Building Energy Efficiency Assessment System and Method Based on Energy Consumption Monitoring

Abstract

This invention relates to the field of energy consumption monitoring technology, specifically to a building energy conservation assessment system and method based on energy consumption monitoring. The system includes an energy consumption sequence construction module, an energy efficiency status quantification module, an abnormal energy consumption element location module, and an energy conservation scheme evaluation module. By collecting building energy consumption data and driving data such as external environment and internal operation, a contextualized data sequence is constructed. A dynamic energy consumption benchmark model is established based on historical data. The energy efficiency health level is calculated and compared with a diagnostic threshold determined through nonparametric kernel density estimation to determine energy consumption anomalies. When anomalies are present, the contribution of each element in the context vector to the energy consumption deviation is quantified to locate the main anomaly source. An energy conservation scheme is generated based on the anomaly elements. The energy conservation effect is predicted through dynamic model simulation and compared with the threshold to evaluate the effectiveness of the scheme, thereby achieving accurate diagnosis and energy conservation optimization of building energy consumption.

Description

Technical Field

[0001] This invention relates to the field of energy consumption monitoring technology, and specifically to a building energy conservation assessment system and method based on energy consumption monitoring. Background Technology

[0002] As the proportion of building energy consumption continues to rise, traditional energy management methods are unable to accurately identify the causes of anomalies and energy-saving potential. Existing technologies mostly rely on static thresholds or single-dimensional analysis, lacking dynamic assessment of the coupled effects of multiple factors. This leads to inaccurate energy efficiency diagnosis and insufficient targeting of energy-saving measures, which restricts the development of building energy efficiency improvement and refined operation and maintenance.

[0003] Existing technologies, such as the invention patent application with publication number CN117421618A, disclose a building energy conservation assessment system and method based on energy consumption monitoring. This includes: acquiring multi-dimensional building energy consumption data for different sampling periods; acquiring the changing trends of the dimensions in different sampling periods; acquiring the correlation degree of a combination of two dimensions; constructing a multi-dimensional sample space; performing hierarchical clustering on all data points in the multi-dimensional sample space by setting different numbers of clusters; obtaining the necessity of merging clusters of data points in the multi-dimensional sample space based on the different hierarchical clustering results and the correlation degree set; obtaining the adjusted merging conditions between any clusters to obtain the adjusted hierarchical clustering results; acquiring abnormal data in the multi-dimensional building energy consumption data based on the adjusted hierarchical clustering results; and predicting abnormal energy consumption patterns of users within the building based on the acquired abnormal data in the multi-dimensional building energy consumption data, thereby making the prediction of building energy consumption more accurate.

[0004] The above solution has at least the following technical problems:

[0005] 1. The above scheme lacks the collection and integrated analysis of key driving factors such as external environmental data and internal operation data of buildings. As a result, the multi-dimensional sample space it constructs is limited to the changing trends and correlations of energy consumption data itself. It cannot establish a dynamic relationship between energy consumption and specific operating scenarios, making it difficult for the model to distinguish between normal energy consumption fluctuations caused by changes in the external environment or internal usage patterns and anomalies caused by actual equipment failures or control failures. This may result in a large number of false alarms or missed alarms, reducing the accuracy of anomaly detection.

[0006] 2. The above-mentioned solutions lack the process of establishing a dynamic energy consumption benchmark model based on historical data, and lack the method of diagnosing the overall energy efficiency status by calculating quantitative indicators such as energy efficiency health. As a result, its anomaly detection relies entirely on the distribution of data points in the cluster space, rather than a relatively reasonable and situation-dependent energy consumption expectation standard. This makes the system only able to identify statistical anomalies that deviate from most data points, but unable to determine whether such deviations are due to substantial energy efficiency degradation or normal, atypical but reasonable operating conditions. For example, high energy consumption under extreme weather conditions may be incorrectly marked as anomaly, while energy efficiency degradation caused by slow equipment performance decline and hidden within the normal data range cannot be effectively captured.

[0007] 3. The above-mentioned solutions lack a mechanism for fine-tuning the root cause of anomalies after they are identified. In particular, they lack an analysis process to assess the contribution of various situational factors to the overall energy consumption deviation. As a result, even if anomalies are detected, it is difficult to quickly and accurately locate the root cause of the problem. The system can only conclude that abnormal data exists, but it cannot clearly point out whether the main cause is an inaccurate reading of the outdoor temperature sensor, a decrease in the efficiency of the chiller unit, or dense human activity. This greatly prolongs the time for fault diagnosis and handling, and affects the efficiency of energy-saving optimization.

[0008] 4. The above-mentioned solution lacks a functional module for automatically generating and pre-evaluating energy-saving solutions based on the located root causes of anomalies. This results in insufficient system closure. While it completes the process from data to anomaly detection and then to pattern prediction, it fails to build the critical link from diagnosis to prescription. The system cannot predict the energy-saving effects of different adjustment strategies through simulation before taking actual action, thus failing to provide decision-makers with quantitative and credible action recommendations. This keeps the entire system at the monitoring and early warning level, making it difficult to directly support the implementation of precise and efficient energy-saving intervention measures, thus limiting its practical value in achieving deep building energy conservation. Summary of the Invention

[0009] To address the aforementioned shortcomings of existing technologies, this invention provides a building energy conservation assessment system and method based on energy consumption monitoring, which can effectively solve the problems of inaccurate energy consumption anomaly detection and lack of targeted energy conservation optimization in the background technology.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a building energy conservation assessment system based on energy consumption monitoring, comprising:

[0011] The energy consumption sequence construction module is used to acquire energy consumption data and drive data in the building to be monitored, and to construct corresponding contextualized data sequences based on the energy consumption data and drive data.

[0012] The energy efficiency status quantification module is used to establish a dynamic energy consumption benchmark model based on contextualized data sequences and various types of energy consumption, and to calculate the energy efficiency health corresponding to various types of energy consumption, thereby assessing whether there are energy consumption anomalies in the building to be monitored.

[0013] The abnormal energy consumption element localization module is used to assess the contribution of each element in the context data sequence based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption when the building under monitoring has abnormal energy consumption, and then locate the element that causes the abnormal energy consumption.

[0014] The energy-saving scheme evaluation module is used to generate corresponding energy-saving schemes based on the elements in the building under monitoring that cause abnormal energy consumption, and then evaluate the effectiveness of the energy-saving schemes.

[0015] Preferably, the specific process of constructing the corresponding contextualized data sequence is as follows: deploy an energy consumption data collector in the building to be monitored, and obtain the energy consumption corresponding to each collection time point based on the set time interval.

[0016] Acquire building external environment data at each collection time point. The building external environment data includes outdoor temperature, outdoor humidity, and solar radiation intensity. The building internal operation data includes key equipment operation status signals and real-time personnel numbers.

[0017] The system collects operational status signals of key equipment at each collection time point, and obtains the real-time number of people in the building to be monitored at each collection time point.

[0018] Using each collection time point as a reference point and combining a fixed collection time interval, each time window is created. Combining the building's external environment data and building's internal operation data corresponding to each collection time point, the representative value corresponding to each time window is calculated, thereby constructing the corresponding contextualized data sequence.

[0019] Preferably, the process of establishing a dynamic energy consumption benchmark model based on various types of energy consumption is as follows: obtaining historical contextualized data sequences, and combining the context vectors in the historical contextualized data sequences to construct dynamic energy consumption benchmark models corresponding to various types of energy consumption.

[0020] Preferably, the specific process for calculating the energy efficiency health level corresponding to various types of energy consumption is as follows: the energy consumption benchmark value corresponding to various types of energy consumption is output through the dynamic energy consumption benchmark model, and the actual energy consumption value and energy consumption benchmark value corresponding to various types of energy consumption in the building to be monitored are combined to calculate the energy efficiency health level corresponding to various types of energy consumption in the building to be monitored.

[0021] Preferably, the process of assessing whether the building to be monitored has abnormal energy consumption is as follows: obtain energy efficiency health data of various types of energy consumption of the building to be monitored within a set historical time period, and obtain historical energy efficiency health datasets of various types of energy consumption.

[0022] Nonparametric kernel density evaluation is performed on historical energy efficiency health datasets of various energy consumption types. Probability density functions corresponding to energy efficiency health for each type of energy consumption are constructed, and inflection points of the probability density functions corresponding to each type of energy consumption are calculated. The energy efficiency health corresponding to each inflection point is the diagnostic threshold for each type of energy consumption.

[0023] By combining the energy efficiency health status corresponding to various types of energy consumption with the corresponding diagnostic thresholds, the building to be monitored is assessed to determine whether there are any abnormal energy consumptions.

[0024] Preferably, the contribution of each element in the assessment scenario data sequence is determined as follows: the time window corresponding to the energy consumption anomaly of the building to be monitored is obtained; based on the type of energy consumption that causes the energy consumption anomaly of the building to be monitored, the energy consumption benchmark value and actual value of the corresponding type of energy consumption are combined at each collection time point; the actual value of the corresponding type of energy consumption at each collection time point is subtracted from the corresponding energy consumption benchmark value to obtain the deviation value corresponding to the type of energy consumption.

[0025] Obtain the representative value of each element corresponding to this type of energy consumption within the time window, as well as the comparative value of each element corresponding to this type of energy consumption in the same historical period. Combine the energy consumption influence factors of each element corresponding to various types of energy consumption to calculate the contribution of each element in the context vector corresponding to this type of energy consumption.

[0026] By combining the deviation value corresponding to this type of energy consumption with the contribution of each element in the scenario vector, the contribution of the benchmark constant term corresponding to this type of energy consumption is calculated.

[0027] By combining the contribution of each element in the context vector corresponding to this type of energy consumption, the percentage contribution of each element in the context vector corresponding to this type of energy consumption is calculated.

[0028] Preferably, the process of locating the element that causes abnormal energy consumption is as follows: based on the contribution percentage of each element in the context vector of the corresponding energy consumption, the contribution percentage of each element is sorted to obtain the element that causes this type of abnormal energy consumption.

[0029] Preferably, the process of generating the corresponding energy-saving scheme is as follows: obtain each physical source associated with the element causing abnormal energy consumption, and calculate the adjustment amount of the control parameters of each physical source associated with the element causing abnormal energy consumption based on the representative value of the current time window of the element causing abnormal energy consumption and the corresponding historical typical operating value.

[0030] By combining the percentage contribution of elements that cause abnormal energy consumption, the actual control parameter adjustment amount of each physical source is calculated.

[0031] Preferably, the process of evaluating the effectiveness of the energy-saving scheme is as follows: Before the energy-saving scheme is implemented, an effect simulation is performed based on the dynamic energy consumption benchmark model. The actual control parameter adjustment amount of each physical source is used as the input condition. The control parameter adjustment amount is converted into a scenario vector and substituted into the dynamic energy consumption benchmark model to output the expected energy consumption of the building to be monitored after the adjustment of each physical source.

[0032] Calculate the expected energy-saving contribution after the simulation of the energy-saving scheme, and combine it with the preset energy-saving effect threshold to evaluate whether the effect of the energy-saving scheme meets the energy-saving requirements.

[0033] In a second aspect, the present invention provides a building energy efficiency assessment method based on energy consumption monitoring, comprising:

[0034] S1. Obtain energy consumption data and drive data within the building to be monitored, and construct corresponding contextualized data sequences based on the energy consumption data and drive data.

[0035] S2. Based on contextualized data sequences, establish dynamic energy consumption benchmark models based on various types of energy consumption, calculate the energy efficiency health corresponding to various types of energy consumption, and then assess whether there are energy consumption anomalies in the buildings to be monitored.

[0036] S3. When an energy consumption anomaly exists in the building to be monitored, the contribution of each element in the situational data sequence is evaluated based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption, thereby locating the element that causes the energy consumption anomaly.

[0037] S4. Based on the elements in the building to be monitored that cause abnormal energy consumption, generate corresponding energy-saving solutions, and then evaluate the effectiveness of the energy-saving solutions.

[0038] The technical solution provided by this invention has the following advantages compared with the known prior art:

[0039] 1. In the process of constructing basic data for energy consumption analysis, this invention integrates external building environment data and internal operating data, and calculates representative values ​​based on time windows to construct contextualized data sequences. This helps to place isolated energy consumption readings in specific operating contexts, laying a solid data foundation for accurately distinguishing between normal energy consumption fluctuations and real anomalies, and avoiding the limitations of relying solely on energy consumption data itself for judgment.

[0040] 2. In establishing energy consumption assessment benchmarks, this invention utilizes historical contextualized data sequences to create dynamic energy consumption benchmark models in the form of linear regression for various types of energy consumption. The benchmark constant term and the energy consumption influencing factors of each context element are automatically fitted and solved using the least squares method. This helps to quantify the specific impact of different driving factors on various types of energy consumption, thereby forming a reasonable expected energy consumption value that dynamically changes with environmental and operational conditions, rather than a fixed threshold, which greatly improves the scientificity and adaptability of energy efficiency assessment.

[0041] 3. In the process of diagnosing the energy efficiency status of a building, the embodiments of the present invention calculate the energy efficiency health level and use a non-parametric kernel density estimation method to automatically determine the diagnostic threshold from historical health data. This is conducive to achieving adaptive and objective anomaly judgment, helps to capture the statistical distribution characteristics of the building's energy efficiency status, accurately identifies energy efficiency degradation phenomena that deviate significantly from historical normal levels, and effectively reduces the risk of misjudgment caused by seasonal and periodic factors.

[0042] 4. In the process of locating the root cause of abnormal energy consumption, the embodiments of the present invention quantify and analyze the contribution of each element in the context data sequence to the overall energy consumption deviation and calculate its contribution percentage, which helps to quickly and accurately identify the main driving factors that cause the abnormality and greatly improves the operation and maintenance efficiency.

[0043] 5. In the process of generating and evaluating energy-saving solutions, this invention associates the located abnormal elements with specific physical sources, calculates the adjustment amount of control parameters, and uses a dynamic energy consumption benchmark model to simulate the effect and predict the expected energy-saving contribution. This is beneficial for quantitatively evaluating the effect of the solution before its actual implementation, ensuring the effectiveness and pertinence of the proposed energy-saving solution, avoiding blind implementation of the transformation, and triggering the solution optimization mechanism to seek a better solution. Ultimately, a complete decision support closed loop is formed from monitoring, diagnosis, location to evaluation and optimization. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0045] Figure 1 This is a schematic diagram of the system structure connection of the present invention.

[0046] Figure 2 This is a schematic diagram of the implementation steps of the present invention. Detailed Implementation

[0047] 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 embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0048] The present invention will be further described below with reference to embodiments.

[0049] Please see Figure 1 As shown, a building energy efficiency assessment system based on energy consumption monitoring includes at least:

[0050] The energy consumption sequence construction module is used to acquire energy consumption data and drive data in the building to be monitored, and to construct corresponding contextualized data sequences based on the energy consumption data and drive data.

[0051] In a specific embodiment, the construction of the corresponding contextualized data sequence is carried out as follows: an energy consumption data collector is deployed in the building to be monitored, and the energy consumption corresponding to each collection time point is obtained from the energy consumption metering point based on the set time interval.

[0052] The building's external environment data for each collection time point is obtained from the API of the local weather station corresponding to the building to be monitored. The building's external environment data includes outdoor temperature, outdoor humidity, and solar radiation intensity.

[0053] Building internal operation data includes key equipment operation status signals and real-time personnel numbers. Operation status signal monitoring instruments are deployed in the key equipment of the building to be monitored to collect the operation status signals of the key equipment at each collection time point.

[0054] By using the access control system of the building to be monitored, the real-time number of people in the building at each data collection point can be obtained.

[0055] Using each data collection time point as a reference point and combining it with a fixed data collection time interval, each time window is created. Combining the building's external environment data and building's internal operation data corresponding to each data collection time point, the representative value corresponding to each time window is calculated.

[0056] By combining the energy consumption at each data collection point with the representative values ​​for each time window, a contextualized data sequence is constructed.

[0057] It should be noted that the energy consumption data acquisition device refers to a conventional intelligent data acquisition device in this field (such as an acquisition terminal based on standard protocols such as Modbus, BACnet, or M-Bus), which is configured to read energy consumption data from various energy consumption metering instruments (such as smart meters, water meters, and gas meters) in the building at set time intervals.

[0058] It should be noted that the set time interval is configured according to the building type and assessment needs. For example, for office buildings with strong energy consumption regularity, a 1-hour interval can be set to achieve macro trend assessment; while for data centers or laboratories with drastic energy consumption fluctuations, an interval of 15 minutes or less needs to be set. This example is only for illustrative purposes and is not the only limitation.

[0059] It should be noted that obtaining data from the local meteorological station API corresponding to the building to be monitored refers to calling the application programming interface provided by the local meteorological department or commercial meteorological service agency; the energy consumption metering points include water meters, electricity meters, and gas meters.

[0060] It should be noted that the key equipment refers to components that have a decisive impact on the total energy consumption of a building, such as chillers and circulating water pumps in central air conditioning systems; the operating status signal monitoring instrument refers to a standard industrial sensor or control module integrated into the controller or power distribution circuit of the above-mentioned equipment, used to collect switch signals (such as start / stop) that can directly reflect the operating mode of the equipment. For example, by monitoring the output signal of the chiller control cabinet, its operating status signals such as cooling / standby and compressor load rate can be obtained.

[0061] It should be noted that, taking the data collection time of 2:00 PM and the data collection interval of 10 minutes as an example, a corresponding 20-minute time window (such as 1:50 to 2:10) is created. The external environmental data and internal operational data of the building corresponding to this data collection time point include continuous data (such as outdoor temperature and outdoor humidity) and event-based data (such as the real-time number of people in the building).

[0062] When calculating the representative value corresponding to the time window, continuous data uses the mean calculation method, such as adding the outdoor temperature values ​​of all collection points within the window and taking the arithmetic mean as the representative temperature of the time window; event-type data takes the maximum value corresponding to each time window, such as taking the highest value of the real-time number of people monitored within the time period as the representative personnel load of the time window.

[0063] It should be noted that the contextualized data sequence is in the form of a time-ordered sequence consisting of {timestamps (each time window), energy consumption, and context vector}, such as {2024-05-20 14:00:00, power consumption 85.6kWh, average outdoor temperature 28.5 degrees Celsius, average solar radiation 450W / m², maximum number of employees on duty 120}}.

[0064] In constructing the basic data for energy consumption analysis, this invention integrates external building environment data and internal operational data, and calculates representative values ​​based on time windows to construct a contextualized data sequence. This helps to place isolated energy consumption readings in specific operational contexts, laying a solid data foundation for accurately distinguishing between normal energy consumption fluctuations and real anomalies, and avoiding the limitations of relying solely on energy consumption data itself for judgment.

[0065] In a specific embodiment, the process of establishing a dynamic energy consumption benchmark model based on various types of energy consumption is as follows: Historical contextualized data sequences are obtained from the database; combined with the context vectors in the historical contextualized data sequences, dynamic energy consumption benchmark models corresponding to various types of energy consumption are constructed. The model form is as follows: ,in Represented as the first This type of energy consumption corresponds to the energy consumption benchmark value. These are the numbers assigned to different types of energy consumption. The value of is a positive integer. This represents the index of each element in the context vector. , The value of is a positive integer. Represented as the first The baseline constant term corresponding to this type of energy consumption, Represented as the first The first type of energy consumption corresponds to the situation vector. Energy consumption influencing factors of each element Represented as the first in the context vector The specific numerical value of the element.

[0066] It should be noted that, taking the outdoor temperature as the first element in the context vector as an example, that is... If it equals 1, then The unit is temperature, such as °C. The unit is kWh / ℃, when When the value equals 2, and the second element in the context vector is a person, then... The unit is a person, at this time The unit is kWh / person.

[0067] It should be noted that the process of setting the energy consumption influence factors of each element in the context vector corresponding to various types of energy consumption is based on the existing multiple linear regression model in mathematical principles, and is obtained by automatically fitting and solving the historical contextualized data sequence through the least squares method. In this scheme, this technical means is applied to the benchmark prediction models established for various types of energy consumption (such as air conditioning electricity and lighting electricity) in the building to be monitored, so that each model obtains the corresponding energy consumption influence factor, thereby quantifying the difference in the impact of the same driving factor (such as outdoor temperature) on different energy consumption. For example, the energy consumption influence factor of outdoor temperature on air conditioning electricity is as high as 4.8 kWh / ℃, while the energy consumption influence factor on lighting electricity is approximately 0.

[0068] It should be noted that, This represents the baseline energy consumption forecast when all driving factors (such as outdoor temperature, number of people, etc.) are zero. Its setting process, along with the energy consumption influencing factors, is automatically determined by fitting historical data using the least squares method. In this scheme... This is used to quantify the inherent base load of various types of energy consumption that is unaffected by changes in external conditions. For example, through fitting, the baseline constant term for air conditioning electricity consumption may be 20 kWh, indicating the fixed energy consumption of systems such as standby and circulating water pumps in the building to be monitored; while the constant term for lighting electricity consumption may be 15 kWh, indicating the base electricity consumption of emergency lighting, etc.; and the constant term for water consumption may be approximately zero.

[0069] In a specific embodiment, the calculation of energy efficiency health corresponding to various types of energy consumption is carried out as follows: Combining the actual energy consumption values ​​and energy consumption benchmark values ​​corresponding to various types of energy consumption in the building to be monitored, the calculation is performed using the following formula: The first [unit / item] inside the building to be monitored was obtained. Energy efficiency and health corresponding to each type of energy consumption , Represented as the first The actual energy consumption value of this type of energy consumption.

[0070] It should be noted that the actual energy consumption value corresponding to each type of energy consumption is the energy consumption of each type of energy consumption at each collection time point in the contextual data sequence; for each data record (i.e., each collection time point) in the contextual data sequence, the above calculation process is performed sequentially.

[0071] In a specific embodiment, the process of assessing whether the building to be monitored has abnormal energy consumption is as follows: obtain energy efficiency health data of various types of energy consumption of the building to be monitored within a set historical time period from the database to obtain historical energy efficiency health datasets of various types of energy consumption.

[0072] Nonparametric kernel density evaluation is performed on historical energy efficiency health datasets of various energy consumption types. Probability density functions corresponding to energy efficiency health for each type of energy consumption are constructed, and inflection points of the probability density functions corresponding to each type of energy consumption are calculated. The energy efficiency health corresponding to each inflection point is the diagnostic threshold for each type of energy consumption.

[0073] The energy efficiency health status corresponding to each type of energy consumption is compared with the corresponding diagnostic threshold. If the energy efficiency health status of a certain type of energy consumption is less than the diagnostic threshold, it is determined that the energy consumption status of that type is abnormal, that is, the building under monitoring has an energy consumption abnormality. Conversely, if the energy efficiency health status of that type of energy consumption is greater than the diagnostic threshold, it is determined that the energy consumption status of that type is not abnormal, that is, the building under monitoring does not have an energy consumption abnormality.

[0074] It should be noted that the logic for setting the historical time period is as follows: it should cover at least two complete typical operating cycles (e.g., two weeks) of the building to be monitored. At the same time, the time period should match the seasonal climate characteristics of the current moment, for example, by selecting the same period in previous years or historical periods with similar climate conditions.

[0075] It should be noted that the nonparametric kernel density assessment is a probability density estimation method based on existing technologies that does not presuppose the data distribution pattern. It transforms discrete data points into continuous probability density curves through smoothing. In this scheme, it is used to assess the distribution patterns of historical health data for various types of energy consumption.

[0076] It should be noted that the specific construction process of the probability density function corresponding to the energy efficiency and health of various types of energy consumption is as follows: Based on the historical energy efficiency and health dataset of various types of energy consumption, a non-parametric kernel density estimation method is used for construction. First, a kernel function (such as the Gaussian kernel function) is selected, and the corresponding bandwidth is determined (used to control the smoothness of the curve, calculated according to the Silverman rule or cross-validation method). Then, a kernel function with the center of the historical energy efficiency and health dataset and the bandwidth is placed on each data point. Finally, these kernel functions are superimposed and summed to form a smooth and continuous probability density function curve.

[0077] It should be noted that the calculation process of the inflection point of the probability density function corresponding to various types of energy consumption is essentially a numerical analysis and function extremum detection method in the existing technology. Specifically, it locates the abrupt change point (i.e., inflection point) of the probability density function curve by calculating the first and second derivatives of the function, which will not be elaborated on further here.

[0078] In establishing energy consumption assessment benchmarks, this invention utilizes historical contextualized data sequences to create dynamic energy consumption benchmark models in the form of linear regression for various types of energy consumption. It then automatically fits and solves the benchmark constant term and the energy consumption influencing factors of each contextual element using the least squares method. This facilitates the quantification of the specific impact of different driving factors on various types of energy consumption, thereby forming a reasonable expected energy consumption value that dynamically changes with environmental and operational conditions, rather than a fixed threshold. This significantly improves the scientific rigor and adaptability of energy efficiency assessments.

[0079] In the process of diagnosing the energy efficiency status of buildings, this invention calculates the energy efficiency health level and uses a non-parametric kernel density estimation method to automatically determine the diagnostic threshold from historical health data. This facilitates the adaptive and objective determination of anomalies, helps to capture the statistical distribution characteristics of building energy efficiency status, accurately identifies energy efficiency degradation phenomena that deviate significantly from historical normal levels, and effectively reduces the risk of misjudgment caused by seasonal and periodic factors.

[0080] The abnormal energy consumption element localization module is used to assess the contribution of each element in the context data sequence based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption when the building under monitoring has abnormal energy consumption, and then locate the element that causes the abnormal energy consumption.

[0081] In a specific embodiment, the contribution of each element in the evaluation scenario data sequence is specifically processed as follows: obtain the time window corresponding to the energy consumption anomaly of the building to be monitored; based on the type of energy consumption that causes the energy consumption anomaly of the building to be monitored, combine the energy consumption benchmark value and actual value of the corresponding type of energy consumption at each collection time point, and subtract the corresponding energy consumption benchmark value from the actual value of the corresponding type of energy consumption at each collection time point to obtain the deviation value corresponding to the type of energy consumption.

[0082] The representative values ​​of each element corresponding to this type of energy consumption within the specified time window are obtained from the contextual data sequence. Comparative values ​​of each element corresponding to this type of energy consumption within the same historical period are retrieved from the database. Combined with the energy consumption impact factors of each element corresponding to various types of energy consumption, the following calculation formula is used: The calculation yields the first element in the scenario vector corresponding to this type of energy consumption. Contribution of each element , and These are respectively represented as the first in the context vector corresponding to this type of energy consumption. The representative value and the comparison value of each element.

[0083] Combining the deviation value corresponding to this type of energy consumption with the contribution of each element in the scenario vector, the calculation formula is as follows: The contribution of the reference constant term corresponding to this type of energy consumption is obtained. , This is expressed as the deviation value corresponding to this type of energy consumption.

[0084] Combining the contribution of each element in the context vector corresponding to this type of energy consumption, the calculation formula is as follows: The first element in the context vector corresponding to this type of energy consumption is calculated. Percentage of contribution of each element ,in This includes the reference constant term and the index of each element. , Represented as the first The contribution amount corresponding to each element.

[0085] It should be noted that the historical period refers to a historical time period that is highly comparable to the current analysis period in terms of season, operating mode, and date type. For example, if the current analysis object is the abnormal air conditioning energy consumption on a certain summer workday, the system will automatically select data from workdays with similar climate conditions in previous summers as the historical period benchmark.

[0086] In a specific embodiment, the process of locating the element that causes abnormal energy consumption is as follows: based on the contribution percentage of each element in the context vector of the corresponding energy consumption, the contribution percentage of each element is sorted, and the element with the highest contribution percentage is selected and determined as the element that causes this type of abnormal energy consumption.

[0087] It should be noted that if the element is outdoor temperature, it indicates that the anomaly is caused by external environmental factors deviating from typical conditions; if the element is a baseline constant, it indicates that the basic energy consumption level of the building under monitoring has undergone a systematic shift, which may be due to equipment performance degradation, control strategy failure, or undetected load changes.

[0088] In locating the root cause of abnormal energy consumption, this invention quantitatively analyzes the contribution of each element in the contextual data sequence to the overall energy consumption deviation and calculates its contribution percentage. This helps to quickly and accurately identify the main driving factors causing the abnormality, greatly improving operation and maintenance efficiency.

[0089] The energy-saving scheme evaluation module is used to generate corresponding energy-saving schemes based on the elements in the building under monitoring that cause abnormal energy consumption, and then evaluate the effectiveness of the energy-saving schemes.

[0090] In a specific embodiment, the generation of the corresponding energy-saving scheme is specifically performed as follows: obtain each physical source associated with the element causing abnormal energy consumption from the database, and calculate the adjustment amount of the control parameters of each physical source associated with the element causing abnormal energy consumption based on the representative value of the current time window of the element causing abnormal energy consumption and the corresponding historical typical operating value.

[0091] Based on the adjustment amount of the control parameters of each physical source associated with the element causing the energy consumption anomaly, and combined with the percentage contribution of the element causing the energy consumption anomaly, the actual adjustment amount of the control parameters of each physical source is obtained.

[0092] It should be noted that the specific meanings of the physical sources associated with the elements causing abnormal energy consumption retrieved from the database are as follows: taking outdoor temperature as an example, the associated physical sources include environmental control equipment such as chillers and fresh air units; when the abnormal element is the real-time number of people, the associated physical sources include equipment linked to personnel activities such as area lighting circuits and fresh air valves.

[0093] It should be noted that the specific process for calculating the control parameter adjustment is as follows: the difference between the current monitoring value of the element causing the energy consumption anomaly and its historical typical operating value (referring to the statistical typical value of the element extracted from historical data under the same season and the same operating mode, such as the mean or median) is used as the initial deviation. This deviation is multiplied by an adjustment coefficient (range 0-1) preset according to the characteristics of the associated physical source equipment to obtain the baseline adjustment range. This baseline adjustment range is then multiplied by the contribution percentage of the element currently causing the energy consumption anomaly (converted to a decimal form of 0-1) to finally generate the actual control parameter adjustment.

[0094] The adjustment coefficient is an empirical value preset based on the characteristics and safe operating range of the physical source equipment. For example, for chiller units with high inertia and slow adjustment, the adjustment coefficient may be set to 0.5 to ensure that the temperature setpoint is adjusted no more than half of the calculated deviation each time; while for lighting circuits with sensitive response, the adjustment coefficient is set to 0.8 to achieve rapid illuminance adjustment.

[0095] In a specific embodiment, the process of evaluating the effectiveness of the energy-saving scheme is as follows: Before the energy-saving scheme is implemented, an effect simulation is performed based on the dynamic energy consumption benchmark model. The actual control parameter adjustment amount of each physical source is used as the input condition. The control parameter adjustment amount is transformed into a scenario vector and substituted into the dynamic energy consumption benchmark model to output the expected energy consumption of the building to be monitored after the adjustment of each physical source.

[0096] Calculate the expected energy-saving contribution after the simulation execution of the energy-saving scheme, compare the expected energy-saving contribution of the energy-saving scheme with the preset energy-saving effect threshold. If the expected contribution is greater than or equal to the energy-saving effect threshold, the energy-saving scheme is determined to meet the energy-saving requirements. If the expected contribution is less than the energy-saving effect threshold, the energy-saving scheme is determined to be unsatisfactory, triggering the scheme optimization mechanism and selecting an alternative energy-saving scheme.

[0097] It should be noted that the calculation method for the expected energy-saving contribution of this energy-saving scheme is the same as the calculation method for the contribution percentage in S3. That is, during the simulation, the energy-saving scheme is executed, and the corresponding parameters are substituted into the formula to calculate the contribution percentage. Therefore, it will not be elaborated further here.

[0098] It should be noted that the specific process of converting control parameter adjustments into scenario vectors is as follows: The system uses a preset control parameter-performance mapping table to convert the control parameter adjustments of the physical source into the corresponding equipment operating energy efficiency change rate. Then, based on this change rate, the corresponding energy consumption impact factor in the dynamic energy consumption benchmark model is synchronously corrected, thereby achieving an equivalent update of the scenario vector. For example, when the system decides to increase the set temperature of the chiller unit by 0.9℃, it queries the mapping table and finds that this operation will increase the unit's energy efficiency ratio by 6%. Therefore, the energy consumption impact factor corresponding to the chiller unit in the dynamic energy consumption benchmark model is synchronously decreased by 6%. Finally, the original scenario data is combined with the corrected dynamic energy consumption benchmark model to predict energy consumption.

[0099] It should be noted that the process for setting the energy-saving effect threshold is the same as the process for setting the diagnostic threshold, and will not be elaborated further here.

[0100] It should be noted that when the energy-saving scheme is deemed ineffective, the triggered scheme optimization mechanism specifically includes: based on the element causing the energy consumption anomaly, recalculating the adjustment amount of the control parameters of each associated physical source (such as increasing the basic adjustment coefficient or adjusting the correction factor); if the adjusted scheme still does not meet the requirements, then matching other strategy templates corresponding to the abnormal element from the strategy knowledge base to generate an alternative scheme; for example, when the expected contribution of the chiller unit set temperature increase strategy is lower than the threshold, then the temperature adjustment range is increased; if it still does not meet the requirements, then switching to the alternative strategy of optimizing the chiller unit operating schedule for re-evaluation.

[0101] Please see Figure 2 As shown, the building energy efficiency assessment method based on energy consumption monitoring includes the following steps:

[0102] S1. Obtain energy consumption data and drive data within the building to be monitored, and construct corresponding contextualized data sequences based on the energy consumption data and drive data.

[0103] S2. Based on contextualized data sequences, establish dynamic energy consumption benchmark models based on various types of energy consumption, calculate the energy efficiency health corresponding to various types of energy consumption, and then assess whether there are energy consumption anomalies in the buildings to be monitored.

[0104] S3. When an energy consumption anomaly exists in the building to be monitored, the contribution of each element in the situational data sequence is evaluated based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption, thereby locating the element that causes the energy consumption anomaly.

[0105] S4. Based on the elements in the building to be monitored that cause abnormal energy consumption, generate corresponding energy-saving solutions, and then evaluate the effectiveness of the energy-saving solutions.

[0106] In the process of generating and evaluating energy-saving solutions, this invention associates the located abnormal elements with specific physical sources, calculates the adjustment amount of control parameters, and uses a dynamic energy consumption benchmark model to simulate the effect and predict the expected energy-saving contribution. This facilitates the quantitative evaluation of the effect before the actual implementation of the solution, ensuring the effectiveness and relevance of the proposed energy-saving solution, avoiding blind implementation of the transformation, and triggering the solution optimization mechanism to seek a better solution. Ultimately, it forms a complete decision support closed loop from monitoring, diagnosis, location to evaluation and optimization.

[0107] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A building energy efficiency assessment system based on energy consumption monitoring, characterized in that, include: The energy consumption sequence construction module is used to acquire energy consumption data and drive data in the building to be monitored, and to construct corresponding contextualized data sequences based on the energy consumption data and drive data. The driving data includes external building environment data and internal building operation data; The energy efficiency status quantification module is used to establish a dynamic energy consumption benchmark model based on contextualized data sequences and various types of energy consumption, and to calculate the energy efficiency health corresponding to various types of energy consumption, thereby assessing whether there are energy consumption anomalies in the building to be monitored. The specific process for assessing whether the building under monitoring has abnormal energy consumption is as follows: Obtain energy efficiency health data for various types of energy consumption of the building to be tested within a set historical time period, and obtain historical energy efficiency health datasets for various types of energy consumption. Nonparametric kernel density evaluation is performed on historical energy efficiency health datasets of various types of energy consumption. Probability density functions of energy efficiency health corresponding to various types of energy consumption are constructed. Inflection points of probability density functions corresponding to various types of energy consumption are calculated. The energy efficiency health corresponding to each inflection point is the diagnostic threshold for various types of energy consumption. By combining the energy efficiency health status corresponding to various types of energy consumption with the corresponding diagnostic thresholds, it is possible to assess whether the building under monitoring has abnormal energy consumption. The abnormal energy consumption element localization module is used to assess the contribution of each element in the situational data sequence based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption when there is an abnormal energy consumption in the building to be monitored, and then locate the element that causes the abnormal energy consumption. The specific process for evaluating the contribution of each element in the contextual data sequence is as follows: Obtain the time window corresponding to the energy consumption anomaly of the building under monitoring. Based on the type of energy consumption that causes the energy consumption anomaly of the building under monitoring, combine the energy consumption benchmark value and actual value of the corresponding type of energy consumption at each collection time point, and subtract the corresponding energy consumption benchmark value from the actual value of the corresponding type of energy consumption at each collection time point to obtain the deviation value corresponding to the type of energy consumption. Obtain the representative value of each element corresponding to this type of energy consumption in the time window, and the comparative value of each element corresponding to this type of energy consumption in the same historical period. Combine the energy consumption influence factors of each element corresponding to each type of energy consumption to calculate the contribution of each element in the context vector corresponding to this type of energy consumption. By combining the deviation value corresponding to this type of energy consumption and the contribution of each element in the scenario vector, the contribution of the baseline constant term corresponding to this type of energy consumption is calculated. By combining the contribution of each element in the context vector corresponding to this type of energy consumption, the percentage contribution of each element in the context vector corresponding to this type of energy consumption is calculated. The energy-saving scheme evaluation module is used to generate corresponding energy-saving schemes based on the elements in the building to be monitored that cause abnormal energy consumption, and then evaluate the effectiveness of the energy-saving schemes. The specific process for generating the corresponding energy-saving solution is as follows: Obtain the physical sources associated with the element causing the energy consumption anomaly, and calculate the control parameter adjustment amount of each physical source associated with the element causing the energy consumption anomaly based on the representative value of the current time window of the element causing the energy consumption anomaly and the corresponding historical typical operating value. By combining the percentage contribution of elements that cause abnormal energy consumption, the actual control parameter adjustment amount of each physical source is calculated.

2. The building energy efficiency assessment system based on energy consumption monitoring according to claim 1, characterized in that, The specific process for constructing the corresponding contextualized data sequence is as follows: Deploy energy consumption data acquisition devices in the building to be monitored, and obtain the energy consumption corresponding to each collection time point based on the set time interval; Acquire building external environment data at each collection time point. The building external environment data includes outdoor temperature, outdoor humidity and solar radiation intensity. The building internal operation data includes key equipment operation status signals and real-time personnel numbers. Collect the operating status signals of key equipment at each collection time point, and obtain the real-time number of people in the building to be monitored at each collection time point; Using each collection time point as a reference point and combining a fixed collection time interval, each time window is created. Combining the building's external environment data and building's internal operation data corresponding to each collection time point, the representative value corresponding to each time window is calculated, thereby constructing the corresponding contextualized data sequence.

3. The building energy efficiency assessment system based on energy consumption monitoring according to claim 2, characterized in that, The specific process for establishing a dynamic energy consumption benchmark model based on various types of energy consumption is as follows: Obtain historical contextualized data sequences and combine them with context vectors to construct dynamic energy consumption benchmark models corresponding to various types of energy consumption.

4. The building energy efficiency assessment system based on energy consumption monitoring according to claim 3, characterized in that, The specific process for calculating the energy efficiency health level corresponding to various types of energy consumption is as follows: By outputting the energy consumption benchmark values ​​corresponding to various types of energy consumption through the dynamic energy consumption benchmark model, and combining the actual energy consumption values ​​and energy consumption benchmark values ​​corresponding to various types of energy consumption in the building to be monitored, the energy efficiency health level corresponding to various types of energy consumption in the building to be monitored is calculated.

5. The building energy efficiency assessment system based on energy consumption monitoring according to claim 4, characterized in that, The process of locating the element that causes abnormal energy consumption is as follows: Based on the percentage contribution of each element in the context vector corresponding to the energy consumption, the percentage contribution of each element is sorted to obtain the element that causes the abnormal energy consumption of that type.

6. The building energy efficiency assessment system based on energy consumption monitoring according to claim 5, characterized in that, The specific process for evaluating the effectiveness of the energy-saving scheme is as follows: Before the energy-saving scheme is implemented, the effect simulation is carried out based on the dynamic energy consumption benchmark model. The actual control parameter adjustment of each physical source is used as the input condition. The control parameter adjustment is transformed into a situation vector and substituted into the dynamic energy consumption benchmark model to output the expected energy consumption of the building to be monitored after the adjustment of each physical source. Calculate the expected energy-saving contribution after the simulation of the energy-saving scheme, and combine it with the preset energy-saving effect threshold to evaluate whether the effect of the energy-saving scheme meets the energy-saving requirements.

7. A building energy efficiency assessment method for implementing the building energy efficiency assessment system based on energy consumption monitoring as described in any one of claims 1-6, characterized in that, Includes the following steps: S1. Obtain energy consumption data and drive data in the building to be monitored, and construct corresponding contextualized data sequences based on the energy consumption data and drive data; S2. Based on contextualized data sequences, establish dynamic energy consumption benchmark models based on various types of energy consumption, calculate the energy efficiency health corresponding to various types of energy consumption, and then assess whether there are energy consumption anomalies in the buildings to be monitored. S3. When an energy consumption anomaly exists in the building to be monitored, the contribution of each element in the situational data sequence is evaluated based on the dynamic energy consumption benchmark model corresponding to various types of energy consumption, thereby locating the element that causes the energy consumption anomaly. S4. Based on the elements in the building to be monitored that cause abnormal energy consumption, generate corresponding energy-saving solutions, and then evaluate the effectiveness of the energy-saving solutions.

Images