A method for energy consumption and energy efficiency analysis for a geothermal energy supply system

Abstract

The present application relates to geothermal energy supply technical field, especially to a kind of energy consumption and energy efficiency analysis method for geothermal energy supply system;By at least one of outdoor air temperature and system load rate is carried out working condition division, construct multidimensional energy efficiency fingerprint library under each working condition, and real-time energy efficiency characteristics and the reference interval of corresponding working condition are compared, the problem that traditional fixed threshold method is difficult to distinguish normal working condition fluctuation and real energy efficiency anomaly is solved, false positive rate and false negative rate of energy efficiency anomaly detection are significantly reduced, can more accurately distinguish normal working condition fluctuation and real energy efficiency anomaly technical problem;Gaussian mixture model and bayesian information criterion are used to automatically determine the optimal working condition division granularity, and the reference interval width of each energy efficiency characteristic dimension under each working condition is adaptively adjusted using sample skewness and kurtosis, so that energy efficiency fingerprint library can truly reflect the skewness and fat tail characteristics of data distribution, improve the statistical rationality of reference interval and the accuracy of anomaly detection.

Description

Technical Field

[0001] This invention relates to the field of geothermal energy supply technology, and in particular to a method for energy consumption and energy efficiency analysis of geothermal energy supply systems. Background Technology

[0002] Energy efficiency analysis of geothermal energy supply systems, especially multi-energy complementary systems such as medium-deep geothermal and gas-fired boilers, is an important means to ensure efficient system operation and reduce carbon emissions. Existing energy efficiency analysis methods typically use the fixed threshold method: for example, setting a fixed lower limit value of the COP of the heat pump unit as the threshold, such as 3.0, or setting a fixed range of the temperature difference between the inlet and outlet water on the buried pipe side as the threshold. When the real-time monitoring indicators exceed the threshold, it is judged as an energy efficiency anomaly.

[0003] However, in actual operation, the energy efficiency index of geothermal energy supply systems is not constant, but fluctuates significantly with changes in operating conditions such as outdoor temperature and system load rate. For example, at the beginning of the heating season, when the outdoor temperature is high and the load rate is low, the COP of the heat pump unit may be at a high level above 5.0; while during the severe cold period, when the outdoor temperature drops sharply and the load rate increases, the COP may drop to around 3.5. Although the latter is a normal performance under this condition, it is very easy to be misjudged as an energy efficiency anomaly under the fixed threshold method. Conversely, if the threshold is set too broadly to cover all operating conditions, when the actual energy efficiency of the system decreases, such as when the performance of the buried pipe heat exchanger deteriorates, it may be missed because the wide threshold has not been reached.

[0004] Therefore, existing energy efficiency analysis methods lack the ability to adaptively distinguish operating conditions. They directly compare energy efficiency indicators under different operating conditions with a fixed reference value, which cannot distinguish between normal operating condition fluctuations and true energy efficiency anomalies. This not only leads to a large number of invalid alarms and increases the burden on operation and maintenance personnel, but may also mask the real trend of energy efficiency degradation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an energy consumption and energy efficiency analysis method for geothermal energy supply systems. This method solves the technical problem that existing energy efficiency analysis methods lack the ability to adaptively distinguish operating conditions, making it difficult to differentiate between normal operating condition fluctuations and actual energy efficiency anomalies.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a method for energy consumption and efficiency analysis of geothermal energy supply systems. This method divides operating conditions based on outdoor temperature and / or system load rate, constructs reference value ranges for energy efficiency characteristics under each operating condition, calculates the deviation between real-time data and the corresponding reference value ranges, and analyzes the causes of anomalies based on the direction of deviation. This effectively reduces false alarm and false negative rates, achieving accurate energy efficiency anomaly localization. The method includes the following steps: S1. Obtain the operating parameter data of the geothermal energy supply system at multiple sampling times. The operating parameter data includes at least: the inlet and outlet water temperature, circulation flow rate, and power data of the buried pipe side and user side of the heat pump unit, as well as the total heat supply and auxiliary heat supply of the system. S2. Based on the aforementioned operating parameter data, construct a multi-dimensional energy efficiency feature vector group; S3. Based on the multi-dimensional energy efficiency feature vector group corresponding to historical operating parameter data, the operating conditions are divided according to at least one of outdoor temperature or system load rate, and the energy efficiency feature reference value range under each operating condition is obtained to construct a multi-dimensional energy efficiency fingerprint database. S4. Extract the real-time multidimensional energy efficiency feature vector at the real-time sampling time, and calculate the deviation between the real-time multidimensional energy efficiency feature vector and the energy efficiency feature reference value range of the current working condition in the multidimensional energy efficiency fingerprint database. S5. In response to a deviation value exceeding a set deviation threshold, analyze the cause of energy efficiency anomalies based on the deviation direction of each energy efficiency characteristic dimension relative to the energy efficiency characteristic reference value range.

[0007] Preferably, in step S2, the multidimensional energy efficiency feature vector set includes at least: The heat exchange temperature difference on the buried pipe side is used to characterize the difference between the inlet and outlet water temperatures on the buried pipe side. The heat pump unit energy efficiency ratio is used to characterize the ratio of the heat output of a heat pump unit to the power of its compressor. The system comprehensive energy efficiency ratio is used to characterize the ratio of the total heat supply of the system to the sum of the power of the heat pump unit compressor and the total electrical power of the circulating pump; User-side supply and return water temperature difference is used to characterize the difference between the user-side supply water temperature and the user-side return water temperature. Auxiliary energy equipment thermal efficiency is used to characterize the ratio of heat supplied by auxiliary energy equipment to energy input of auxiliary energy equipment.

[0008] Preferably, in step S3, the working condition division adopts a Gaussian mixture model as the clustering algorithm, and the Bayesian information criterion is used as the optimization objective for selecting the number of clusters. The Gaussian mixture model is fitted under different candidate values ​​of the number of clusters, and the number of clusters that maximizes the value of the Bayesian information criterion is selected as the optimal number of clusters for the working condition division. The expression for the Bayesian information criterion is: ; ; In the above formula, This indicates that when the number of clusters in the Gaussian mixture model is... The Bayesian information criterion value at time oscillation comprehensively measures the model's goodness of fit to the data and the model's complexity, by maximizing... To select the optimal number of clusters, a balance must be struck between avoiding overfitting and underfitting. For Gaussian mixture models, the parameter estimates are... The following dataset The maximum likelihood function value, This represents the number of Gaussian components in the Gaussian mixture model, i.e., the number of clusters. The number of samples for historical operating parameter data. The mixing weights for the k-th Gaussian component. Let be the mean vector of the k-th Gaussian component. Let be the covariance matrix of the k-th Gaussian component. Let the probability density function of the k-th Gaussian component be denoted as . For the multidimensional energy efficiency feature vector of the nth sample, maximize... A balance is struck between avoiding overfitting and underfitting, and the granularity of the operating conditions is obtained to match the actual operating characteristics of the geothermal energy supply system.

[0009] Preferably, obtaining the reference value range of energy efficiency characteristics under various operating conditions specifically includes: For any working condition category and their corresponding energy efficiency characteristic dimensions Obtain the first of the historical multidimensional energy efficiency feature vectors under this operating condition. Datasets with dimensional features ,in The sample size under this operating condition is used to calculate the mean of this energy efficiency characteristic dimension under this operating condition. and standard deviation ; With the mean Centered on the standard deviation preset multiple The reference value interval is constructed using half the width of the reference value interval. ,in, The calculation formula is based on the preset benchmark multiple and the sample skewness of the corresponding dimension feature data. ; ; ; In the above formula, As a base multiple, the range of values ​​is , Indicates the type of working condition Next Sample skewness of dimensional feature data Indicates the type of working condition Next Sample kurtosis of dimensional feature data This is an adjustment coefficient, and its value ranges from 0 to 1. It is an infinitesimal positive number, used to avoid the denominator being zero. Indicates the first The value of the i-th sample in the dataset with dimensional features.

[0010] Preferably, the deviation value is calculated using Mahalanobis distance, and the specific steps are as follows: For the real-time sampling time t, let the extracted real-time multidimensional energy efficiency feature vector be... ,in, This indicates that at real-time sampling time t, the energy efficiency feature dimension is... The parameter value at that time, This represents the total number of dimensions of energy efficiency features; Let the reference value range in the multidimensional energy efficiency fingerprint database that matches real-time operating conditions be composed of the mean vectors of each dimension. Covariance Matrix Characterization; the mean vector The covariance matrix is ​​obtained by calculating the mean of each dimension. It can be calculated using the following formula: ; In the above formula, For working conditions The number of historical samples below Indicates working conditions The i-th historical multidimensional energy efficiency feature vector; Based on the mean vector Covariance Matrix and real-time multidimensional energy efficiency feature vector Calculate the Mahalanobis distance as the deviation value. The calculation formula is as follows: ; Preferably, in step S5, the analysis of the causes of energy efficiency anomalies specifically includes: S51. A mapping rule base between deviation direction patterns and abnormal causes is pre-built and stored. Each rule in the mapping rule base includes a condition part and a conclusion part. The condition part is a combination pattern of deviation direction of several energy efficiency feature dimensions. The conclusion part is the abnormal cause corresponding to the combination pattern. S52. Based on the deviation direction of each energy efficiency feature dimension relative to the energy efficiency feature reference value range determined in real time, construct a real-time deviation direction pattern vector. ,in, For the first The deviation direction state value for each energy efficiency characteristic dimension is determined based on the corresponding reference value range, as follows: ; in, For real-time sampling time t, the first... The values ​​of each energy efficiency feature dimension For this feature dimension in the working condition The mean of the following; S53, the real-time deviation direction mode vector Perform pattern matching with the condition part of each rule in the mapping rule base, and calculate the matching degree; S54. The abnormal cause corresponding to the rule with the highest matching degree is used as the inferred cause of the real-time energy efficiency anomaly, and the energy efficiency anomaly diagnosis result is output; when the matching degree of the rule with the highest matching degree is lower than the preset matching degree threshold, the real-time deviation direction mode vector is marked as a new mode to update the mapping rule library.

[0011] The present invention also provides an energy consumption and energy efficiency analysis system for geothermal energy supply systems, including a processor and a memory, wherein the memory is used to store a computer program, and the computer program is executed by the processor to implement the energy consumption and energy efficiency analysis method.

[0012] By employing the above technical solution, the present invention provides a method for energy consumption and energy efficiency analysis of geothermal energy supply systems, which has at least the following beneficial effects: 1. This invention divides operating conditions based on at least one of outdoor temperature and system load rate, constructs a multi-dimensional energy efficiency fingerprint database for each operating condition, and compares the deviation between real-time energy efficiency characteristics and the reference range of the corresponding operating condition. This solves the problem that the traditional fixed threshold method is difficult to distinguish between normal operating condition fluctuations and real energy efficiency anomalies, and significantly reduces the false alarm rate and false negative rate of energy efficiency anomaly detection.

[0013] 2. This invention uses a Gaussian mixture model and Bayesian information criterion to automatically determine the optimal granularity of working conditions, and uses sample skewness and kurtosis to adaptively adjust the reference interval width of each energy efficiency feature dimension under each working condition, so that the energy efficiency fingerprint database can truly reflect the skewness and heavy tails of the data distribution, thereby improving the statistical rationality of the reference interval and the accuracy of anomaly detection.

[0014] 3. This invention comprehensively evaluates the deviation between the multidimensional energy efficiency feature vector and the fingerprint database reference center by Mahalanobis distance, eliminating the dimensional differences and correlation effects between the feature dimensions, and realizes automatic inference of the cause of anomalies based on deviation direction pattern matching, thus elevating energy efficiency analysis from whether it is abnormal to the diagnostic level of what the cause is, providing operation and maintenance personnel with an actionable decision-making basis. Attached Figure Description

[0015] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the energy consumption and energy efficiency analysis method for geothermal energy supply systems according to the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0017] To address the technical problem that existing energy efficiency analysis methods lack the ability to adaptively differentiate operating conditions, making it difficult to distinguish between normal operating fluctuations and true energy efficiency anomalies, this implementation provides an energy consumption and energy efficiency analysis method for geothermal energy supply systems, including the following steps: S1. The energy efficiency analysis of the geothermal energy supply system relies on a quantitative description of the actual operating status of the system. Therefore, it is necessary to collect raw monitoring data that can reflect the energy conversion and transfer process of each key link of the system. Therefore, in this embodiment, by deploying metering devices such as temperature sensors, flow meters and electricity meters on the buried pipe side of the heat pump unit and the user side, the operating parameter data of the geothermal energy supply system at multiple sampling times are obtained. Multiple sampling times constitute a time series, and the sampling interval can be set according to the dynamic characteristics of the system, such as 1 minute to 15 minutes, in order to capture the trend of energy efficiency changes. The operating parameter data should include at least the inlet and outlet water temperatures, circulation flow rates, and electrical power data of the heat pump unit's buried pipe side and the user side, as well as the total system heating capacity and auxiliary energy heating capacity. Among them, the main data obtained from the heat pump unit's buried pipe side include inlet water temperature, outlet water temperature, and its circulation flow rate, etc. The main data obtained from the heat pump unit's user side include supply water temperature, return water temperature, and its circulation flow rate, etc. Electrical power data includes the power of the heat pump unit's compressor and the total electrical power of the circulation pump, etc. The total system heating capacity can be calculated from the temperature difference and flow rate of the supply and return water in the user-side water system, or directly read from the heat meter, etc. Auxiliary energy heating capacity includes the instantaneous heating capacity of the gas boiler, as well as its energy input, etc.

[0018] S2. Due to the different dimensions and dispersed physical meanings of the original operating parameter data, they cannot be directly used for comprehensive analysis of energy efficiency anomalies. They need to be transformed into standardized energy efficiency characteristic indicators that reflect the energy conversion efficiency of each link in the system. Furthermore, a single energy efficiency indicator (such as monitoring only the COP of a heat pump unit) cannot cover all aspects, making it difficult to pinpoint the root cause when energy efficiency anomalies occur. Therefore, this embodiment, based on operating parameter data and according to basic thermodynamic principles, extracts energy efficiency characteristic indicators with clear physical meanings from the original operating parameter data. Each indicator corresponds to a key energy conversion or transfer link in the system, and a multi-dimensional energy efficiency characteristic vector group is constructed, which includes at least the following five dimensions: The temperature difference of heat exchange on the buried pipe side The temperature difference between the inlet and outlet water temperatures on the buried pipe side is calculated to characterize the ability of the buried pipe heat exchanger to extract heat from the soil and rock. When the buried pipe extracts heat from the soil and rock, the fluid temperature decreases. The larger the temperature difference on the buried pipe side, the greater the heat extraction. However, an excessively large temperature difference may indicate insufficient circulation flow. An excessively small temperature difference on the buried pipe side may indicate scaling of the buried pipe heat exchanger or thermal attenuation of the soil and rock. This indicator is a direct reflection of the heat exchange capacity of the buried pipe side.

[0019] Energy efficiency ratio of heat pump units EER (Energy Efficiency Ratio) is calculated by the ratio of the heat output of a heat pump unit to the power of its compressor. It reflects the energy conversion efficiency of the heat pump unit itself and is the most crucial indicator for evaluating heat pump performance. When compressor efficiency decreases, refrigerant leaks, or heat exchanger fouling increases, EER will decrease significantly. System overall energy efficiency ratio The efficiency is calculated by dividing the total heat output of the system by the sum of the power of the heat pump unit's compressor and the total electrical power of the circulating pump. This factor incorporates the energy consumption of the circulating pump, reflecting the overall efficiency of the entire heat pump power supply system. When the efficiency of the distribution system decreases, such as when the water pump deviates from its high-efficiency zone or when pipe resistance increases, this factor is applied. The decrease will be greater than This helps to distinguish between host problems and distribution problems; User-side supply and return water temperature difference This is obtained by calculating the difference between the user-side supply water temperature and the user-side return water temperature, reflecting the degree of matching between the user-side load and the transmission and distribution system. An excessively high flow rate may indicate insufficient flow at the point of sale, such as a valve not being fully open, a clogged filter, or an airlock. A value that is too low may indicate that the flow rate is too high or the building load is reduced. This indicator is used to diagnose the hydraulic balance and end-point heat exchange status on the user side.

[0020] Auxiliary power equipment thermal efficiency This is obtained by calculating the ratio of heat supplied by auxiliary power equipment to energy input of auxiliary power equipment. For gas-fired boilers, It can be converted into power corresponding to gas consumption and is used to measure the energy conversion efficiency of auxiliary energy equipment (such as gas boilers). This indicator is used to identify the efficiency decline of the auxiliary energy equipment itself, such as burner aging, air-fuel ratio imbalance and flue gas temperature increase.

[0021] The parameters of the above five dimensions can be used as the five parameters of a vector to construct a multidimensional energy efficiency feature vector group.

[0022] S3. Considering that the normal range of energy efficiency characteristic indicators varies greatly under different outdoor temperatures or different system load rates, using a globally unified reference interval may lead to a large number of false alarms or missed alarms. Therefore, it is necessary to divide the historical operating data according to the operating conditions so that the data under each operating condition has similar operating characteristics, thereby obtaining the statistical distribution of energy efficiency characteristic indicators under that operating condition. Therefore, this embodiment is based on the multi-dimensional energy efficiency characteristic vector group corresponding to the historical operating parameter data, and divides the operating conditions according to at least one of the outdoor temperature or system load rate, dividing the data into several operating condition categories. Then, within each operating condition category, the reference value interval of each energy efficiency characteristic dimension is calculated to obtain the energy efficiency characteristic reference value interval under each operating condition, so as to construct a multi-dimensional energy efficiency fingerprint library. The multi-dimensional energy efficiency fingerprint library, that is, given the operating condition category, can output the normal fluctuation range of each energy efficiency characteristic indicator under that operating condition, such as the normal fluctuation range of each characteristic dimension parameter in the multi-dimensional energy efficiency characteristic vector group. The quality of operating condition segmentation directly affects the accuracy of the energy efficiency fingerprint database. Traditional operating condition segmentation often relies on manual experience, such as dividing outdoor temperature into several fixed intervals. However, the energy efficiency characteristics of geothermal energy supply systems are often not uniformly segmented, and the boundaries between different operating conditions are blurred. Fixed interval segmentation may be difficult to adapt to the climate characteristics and system characteristics of different regions, easily resulting in excessively large or small variances in the data distribution within the same operating condition. Therefore, in this embodiment, the operating condition segmentation adopts a Gaussian Mixture Model (GMM) as the clustering algorithm, and the Bayesian Information Criterion (BIC) is used as the optimization objective for selecting the number of clusters. The GMM is fitted under different candidate values ​​for the number of clusters, and the number of clusters that maximizes the value of the Bayesian Information Criterion is selected as the optimal number of clusters for operating condition segmentation, which is more flexible than spherical clustering such as K-means. The specific use of the Gaussian Mixture Model and the Bayesian Information Criterion will be introduced below: For Gaussian mixture models, it is assumed that the historical multidimensional energy efficiency feature vector Depend on The probability density function of the mixture of Gaussian distributions, where each Gaussian component represents a working condition category, is as follows: ; in, Let be the mixing weight of the k-th component. The sum of the mixing weights of all components is 1, and all are greater than 0. Let represent the mean vector of the k-th component. Let the covariance matrix of the k-th component be represented. This represents a multivariate Gaussian distribution. By using the Expectation-Maximization (EM) algorithm to estimate the parameters of historical data, the parameters of each component can be obtained. , and Then, based on the posterior probability, each sample is assigned to the component with the highest probability, thus completing the work condition classification.

[0023] For the Bayesian information criterion, due to the number of clusters... It needs to be determined in advance that this embodiment uses BIC to automatically select the optimal number of clusters. To avoid relying on manual guessing, its expression can be represented as: ; ; In the above formula, This indicates that when the number of clusters in the Gaussian mixture model is... The Bayesian information criterion value at time oscillation comprehensively measures the model's goodness of fit to the data and the model's complexity, by maximizing... To select the optimal number of clusters, a balance must be struck between avoiding overfitting and underfitting. For Gaussian mixture models, the parameter estimates are... The following dataset The maximum likelihood function value, This represents the number of Gaussian components in the Gaussian mixture model, i.e., the number of clusters. The number of samples for historical operating parameter data. The mixing weights for the k-th Gaussian component. Let be the mean vector of the k-th Gaussian component. Let be the covariance matrix of the k-th Gaussian component. Let the probability density function of the k-th Gaussian component be denoted as . Let n be the multidimensional energy efficiency feature vector of the nth sample. This term measures the goodness of fit of the model to the data; the smaller the value, the better. Used to penalize the complexity of the model. The bigger the punishment, the greater the penalty, through different Fit GMM to values ​​(e.g., K=2 to K=10) and calculate the corresponding values. Value, select to make The largest value As the optimal number of clusters.

[0024] For example, historical operational data from at least one complete heating (or cooling) season is collected and processed through steps S1 and S2 to obtain a multidimensional energy efficiency feature vector set. Then, a range for the number of candidate clusters is set, such as the interval [2, 10]. For each... The GMM is fitted using the EM algorithm, and the maximum likelihood value is calculated. , and then calculate Select to make The largest value Then the The value was used as the final cluster number to refit the GMM, and the results for each working condition were obtained. , and Parameters such as these are used to assign each historical data sample to the corresponding operating condition.

[0025] Traditional methods construct reference ranges for energy efficiency characteristics using the mean as the center and a fixed multiple of the standard deviation as the range, such as μ±2σ or μ±3σ. However, actual data often does not conform to a normal distribution. When the data exhibits skewness or excess kurtosis, the fixed-width range may be too wide (missing true anomalies) or too narrow (falsely reporting normal fluctuations). This embodiment obtains the reference ranges for energy efficiency characteristics under various operating conditions through the following steps: For any working condition category and their corresponding energy efficiency characteristic dimensions Obtain the first of the historical multidimensional energy efficiency feature vectors under this operating condition. Datasets with dimensional features ,in The sample size for this operating condition is given, and the mean and standard deviation of this dataset are calculated to obtain the mean of this energy efficiency feature dimension under this operating condition. and standard deviation .

[0026] With this mean Centered on this standard deviation preset multiple As half the width of the reference value range, Also known as kurtosis, it is used to construct a reference value range, and its expression can be represented as: ,in, The calculation formula is based on the preset benchmark multiple and the sample skewness of the corresponding dimension feature data. ; ; ; In the above formula, As a base multiple, the range of values ​​is If the data is normally distributed, the corresponding confidence interval is approximately 95% to 99%. Indicates the type of working condition Next The sample skewness of dimensional feature data is used to represent the asymmetry of the distribution. When the vertical value is 0, the distribution is symmetrical; when it is greater than 0, it is right-skewed (i.e., the long tail is on the right); and when it is less than 0, it is left-skewed. Indicates the type of working condition Next The kurtosis of dimensional feature data represents the thickness of the tails of the distribution relative to a normal distribution. When it is greater than 0, it indicates a leptokurtic distribution with many extreme values; when it is less than 0, it indicates a flattened distribution with few extreme values. This is an adjustment coefficient, ranging from 0 to 1, used to control the adaptive strength; a value of 0.5 is typically used. It is an infinitesimal positive number, used to avoid the denominator being zero. Indicates the first The value of the i-th sample in a dataset with dimensional features. When the data distribution is close to normal, and All are close to 0, at this time Approximately equal to When the intervals are compact, and the data exhibits significant skewness or thick tails, then... or As the absolute value increases, at this time The reference range is automatically increased to accommodate normal fluctuations and avoid misjudging normal values ​​as abnormal.

[0027] S4. During real-time system operation, it is necessary to quantify the difference between the current energy efficiency status and the normal status in the multi-dimensional energy efficiency fingerprint database to determine whether there is an energy efficiency anomaly. Therefore, it is first necessary to determine the current operating condition category based on at least one of the current outdoor temperature or system load rate. Simultaneously, the real-time multi-dimensional energy efficiency feature vector at the real-time sampling time is extracted, and the deviation between this real-time multi-dimensional energy efficiency feature vector and the energy efficiency feature reference value range of the current operating condition in the multi-dimensional energy efficiency fingerprint database is calculated. Simple deviation calculation methods, such as calculating the absolute values ​​of deviations independently for each dimension and summing them, or using Euclidean distance, have two drawbacks: first, the dimensions of each feature dimension are different; for example, directly adding temperature, ratio, and efficiency lacks physical meaning; second, the correlation between dimensions is ignored. For example, the heat exchange temperature difference on the buried pipe side is often positively correlated with the energy efficiency ratio of the heat pump unit. Mahalanobis distance, through the inverse matrix of the covariance matrix, performs a whitening transformation on the feature space, which can eliminate the influence of dimensions and correlations, making the deviation values ​​under different operating conditions comparable. Therefore, in this embodiment, the deviation value is calculated using Mahalanobis distance, and the specific steps are as follows: For the real-time sampling time t, let the extracted real-time multidimensional energy efficiency feature vector be... ,in, This indicates that at real-time sampling time t, the energy efficiency feature dimension is... The parameter value at that time, This represents the total number of dimensions of energy efficiency features; Suppose that the reference value range in the multidimensional energy efficiency fingerprint database that matches real-time operating conditions is composed of the mean vectors of each dimension. Covariance Matrix Representation; mean vector Covariance Matrix To obtain the current operating conditions from the multi-dimensional energy efficiency fingerprint database The mean vector and covariance matrix are given below. Based on historical samples under this operating condition, the following formula is used for calculation: ; In the above formula, For working conditions The number of historical samples below Indicates working conditions The i-th historical multidimensional energy efficiency feature vector; Based on the mean vector Covariance Matrix and real-time multidimensional energy efficiency feature vector Calculate the Mahalanobis distance as the deviation value. The calculation formula is as follows: ; S5. In response to a deviation value exceeding a set deviation threshold, which can be determined based on the percentile of the Mahalanobis distance distribution of historical normal data (e.g., 95% or 99%), the cause of the energy efficiency anomaly is analyzed based on the direction of deviation of each energy efficiency characteristic dimension relative to the energy efficiency characteristic reference value range. Otherwise, if the deviation value is less than or equal to the set deviation threshold, the system is considered normal. Knowing only that the system's energy efficiency is abnormal is insufficient to guide maintenance personnel in taking effective measures. It is necessary to pinpoint the specific link where the anomaly occurs, such as the buried pipe side, heat pump unit, user side, and auxiliary energy equipment, as well as the possible types of faults. Different causes of anomalies can lead to deviations in each energy efficiency characteristic dimension in specific combinations (e.g., insufficient flow on the buried pipe side can lead to...). negative deviation Positive deviation, and (Possibly normal). By establishing a rule base mapping deviation patterns to abnormal causes, rapid and automated cause inference can be achieved. Specifically, in this embodiment, the specific steps for analyzing the causes of energy efficiency anomalies include: S51. Pre-build and store a mapping rule base between deviation direction patterns and abnormal causes. The mapping rule base can be pre-built based on domain expert knowledge, equipment manufacturer manuals and historical failure cases. Each rule in the mapping rule base includes a condition part and a conclusion part. The condition part is a combination pattern of deviation direction of several energy efficiency feature dimensions, and the conclusion part is the abnormal cause corresponding to the combination pattern.

[0028] S52. Based on the deviation direction of each energy efficiency feature dimension relative to the energy efficiency feature reference value range determined in real time, construct a real-time deviation direction pattern vector. ,in, For the first The deviation direction state value for each energy efficiency characteristic dimension is determined based on the corresponding reference value range, as follows: ; in, For real-time sampling time t, the first... The values ​​of each energy efficiency feature dimension For this feature dimension in the working condition The mean of the vector is given, and each component of the vector takes only +1, -1, or 0, representing a positive deviation, a negative deviation, or being within the normal range for that energy efficiency feature dimension, respectively. Taking a vector with 5 parametric features as an example, Table 1 below shows a partial mapping rule table: Table 1 Partial Mapping Rules Table

[0029] In the table above, "-" indicates that the temperature is not included in the matching and is a wildcard.

[0030] S53, the real-time deviation direction mode vector Pattern matching is performed with the conditional parts of each rule in the mapping rule base, and the matching degree is calculated. The matching degree can be obtained by statistically analyzing the real-time deviation direction pattern vector. The number of features consistent with the deviation patterns corresponding to each rule in the mapping rule base is obtained; for example, if the real-time deviation direction pattern vector... If the pattern is (-1,0,0,+1,-1) and the deviation pattern of rule R1 is (-1,0,0,+1,-), then the first four features of the two are the same, and their matching degree can be determined as 4. Alternatively, the matching degree can be determined as 4 / 5 by dividing the number of matching feature dimensions by the total number of feature dimensions.

[0031] S54. The abnormal cause corresponding to the rule with the highest matching degree is used as the inferred cause of the real-time energy efficiency anomaly, and the energy efficiency anomaly diagnosis result is output. When the matching degree of the rule with the highest matching degree is lower than the preset matching degree threshold, the real-time deviation direction mode vector is marked as a new mode to update the mapping rule base, so as to realize the adaptive expansion and continuous improvement of the mapping rule base.

[0032] This embodiment also provides an energy consumption and energy efficiency analysis system, including a processor and a memory. The memory is used to store computer programs, and when the computer programs are executed by the processor, they implement energy consumption and energy efficiency analysis methods.

[0033] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0034] The following scenario compares the differences between this method and the traditional fixed threshold method: A medium-deep geothermal energy system operates during the heating season, with outdoor temperatures varying between -5℃ and 10℃. Using the traditional fixed threshold method, the lower limit of the heat pump unit's COP is set at 3.5. In actual operation, the following normal operating condition fluctuations and actual energy efficiency anomalies occur: Normal operating condition fluctuations: When the outdoor temperature drops to -5℃ and the system load rate rises to 90%, the COP of the heat pump unit drops to 3.4 due to the decrease in evaporation temperature, which is lower than the fixed threshold of 3.5, triggering an energy efficiency anomaly alarm. However, after on-site verification, all equipment in the system is operating normally, and this decrease in COP is a normal phenomenon under this low temperature and high load condition; the alarm was a false alarm.

[0035] Actual energy efficiency anomaly: When the outdoor temperature is 5℃ and the load rate is 50%, the COP, which should be in the high-efficiency range, should be above 4.8. However, due to the decrease in soil temperature around the buried pipe, the actual COP drops to 4.0, which is still higher than the fixed threshold of 3.5. Therefore, this anomaly is not reported.

[0036] In the method of this embodiment, in step S3, based on historical normal operation data (covering the entire heating season), the operating conditions are divided according to outdoor temperature and system load rate. Using Gaussian mixture model clustering, the data is automatically divided into three typical operating conditions: Condition A: Temperature ≤ 0℃, Load rate ≥ 80%; Condition B: 0℃ < Temperature ≤ 8℃, 40% ≤ Load rate < 80%; Condition C: Temperature > 8℃, Load rate < 40%. Reference value ranges for each energy efficiency characteristic are calculated for each operating condition. For example, the reference range for COP under Condition A is [3.2, 3.7]; the reference range for COP under Condition C is [4.5, 5.2]. Then, through steps S4 and S5, for the above-mentioned normal fluctuation situation: that is, when the outdoor temperature is -5℃ and the load rate is 90%, the real-time COP = 3.4 falls within the reference range [3.2, 3.7] of Condition A, with all dimensions deviating from the threshold. The calculated Mahalanobis distance is less than the threshold, and no abnormal alarm is triggered, successfully avoiding false alarms. In response to the aforementioned real-world anomaly, when the outdoor temperature is 5℃ and the load rate is 50%, the system identifies the current operating condition as B, with a COP reference range of [4.0, 4.8]. The measured COP of 4.0 has reached the lower limit of this range. Simultaneously, the heat exchange temperature difference on the buried pipe side shows a negative deviation (-1), with the deviation direction pattern vector being (-1, -1, -1, 0, -). This pattern matches the "soil temperature decay around buried pipe" rule in the rule library, with a matching degree greater than the threshold. The system outputs the diagnostic analysis result: "The cause of the energy efficiency anomaly is the decay of soil temperature around the buried pipe. It is recommended to check the historical cumulative heat extraction and evaluate soil restoration measures," achieving accurate alarm and cause location. The table below, based on the heating season operation data of the medium-deep geothermal energy supply system, presents the results of a quantitative evaluation of the performance of the traditional fixed threshold method and the method of this invention in energy efficiency anomaly detection:

[0037] Therefore, the method of the present invention is significantly superior to the traditional fixed threshold method in terms of false alarm rate, false negative rate, diagnostic accuracy and adaptability to operating conditions, and can output specific causes of anomalies, providing effective support for operation and maintenance decisions.

[0038] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0039] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for energy consumption and energy efficiency analysis of geothermal energy supply systems, characterized in that, Includes the following steps: S1. Obtain the operating parameter data of the geothermal energy supply system at multiple sampling times. The operating parameter data includes at least: the inlet and outlet water temperature, circulation flow rate, and power data of the buried pipe side and user side of the heat pump unit, as well as the total heat supply and auxiliary heat supply of the system. S2. Based on the aforementioned operating parameter data, construct a multi-dimensional energy efficiency feature vector group; S3. Based on the multi-dimensional energy efficiency feature vector group corresponding to historical operating parameter data, the operating conditions are divided according to at least one of outdoor temperature or system load rate, and the energy efficiency feature reference value range under each operating condition is obtained to construct a multi-dimensional energy efficiency fingerprint database. S4. Extract the real-time multidimensional energy efficiency feature vector at the real-time sampling time, and calculate the deviation between the real-time multidimensional energy efficiency feature vector and the energy efficiency feature reference value range of the current working condition in the multidimensional energy efficiency fingerprint database. S5. In response to a deviation value exceeding a set deviation threshold, analyze the cause of energy efficiency anomalies based on the deviation direction of each energy efficiency characteristic dimension relative to the energy efficiency characteristic reference value range.

2. The energy consumption and energy efficiency analysis method according to claim 1, characterized in that, In step S2, the multidimensional energy efficiency feature vector set includes at least: The heat exchange temperature difference on the buried pipe side is used to characterize the difference between the inlet and outlet water temperatures on the buried pipe side. The heat pump unit energy efficiency ratio is used to characterize the ratio of the heat output of a heat pump unit to the power of its compressor. The system comprehensive energy efficiency ratio is used to characterize the ratio of the total heat supply of the system to the sum of the power of the heat pump unit compressor and the total electrical power of the circulating pump; User-side supply and return water temperature difference is used to characterize the difference between the user-side supply water temperature and the user-side return water temperature. Auxiliary energy equipment thermal efficiency is used to characterize the ratio of heat supplied by auxiliary energy equipment to energy input of auxiliary energy equipment.

3. The energy consumption and energy efficiency analysis method according to claim 1, characterized in that, In step S3, the working condition division adopts a Gaussian mixture model as the clustering algorithm, and the Bayesian information criterion is used as the optimization objective for selecting the number of clusters. The Gaussian mixture model is fitted under different candidate values ​​of the number of clusters, and the number of clusters that maximizes the value of the Bayesian information criterion is selected as the optimal number of clusters for the working condition division.

4. The energy consumption and energy efficiency analysis method according to claim 1, characterized in that, In step S3, obtaining the reference value range of energy efficiency characteristics under each operating condition specifically includes: For any working condition category and their corresponding energy efficiency characteristic dimensions Obtain the first of the historical multidimensional energy efficiency feature vectors under this operating condition. Datasets with dimensional features ,in This represents the number of samples under this operating condition. Calculate the mean and standard deviation of each parameter in the dataset to obtain the mean of this energy efficiency characteristic dimension under the operating conditions. and standard deviation ; With the mean Centered on the standard deviation preset multiple The reference value interval is constructed using half the width of the reference value interval. ,in, The calculation formula is based on the preset benchmark multiple and the sample skewness of the corresponding dimension feature data. ； ； ； In the above formula, As a base multiple, the range of values ​​is , Indicates the type of working condition Next Sample skewness of dimensional feature data Indicates the type of working condition Next Sample kurtosis of dimensional feature data This is an adjustment coefficient, and its value ranges from 0 to 1. It is an infinitesimal positive number, used to avoid the denominator being zero. Indicates the first The value of the i-th sample in the dataset with dimensional features.

5. The energy consumption and energy efficiency analysis method according to claim 1, characterized in that, In step S4, the deviation value is calculated using Mahalanobis distance, and the specific steps are as follows: For the real-time sampling time t, let the extracted real-time multidimensional energy efficiency feature vector be... ,in, This indicates that at real-time sampling time t, the energy efficiency feature dimension is... The parameter value at that time, This represents the total number of dimensions of energy efficiency features; Let the reference value range in the multidimensional energy efficiency fingerprint database that matches real-time operating conditions be composed of the mean vectors of each dimension. Covariance Matrix Characterization; Based on the mean vector Covariance Matrix and real-time multidimensional energy efficiency feature vector Calculate the Mahalanobis distance as the deviation value. .

6. The energy consumption and energy efficiency analysis method according to claim 1, characterized in that, In step S5, the analysis of the causes of energy efficiency anomalies specifically includes: S51. A mapping rule base between deviation direction patterns and abnormal causes is pre-built and stored. Each rule in the mapping rule base includes a condition part and a conclusion part. The condition part is a combination pattern of deviation direction of several energy efficiency feature dimensions. The conclusion part is the abnormal cause corresponding to the combination pattern. S52. Construct a real-time deviation direction pattern vector based on the deviation direction of each energy efficiency feature dimension relative to the energy efficiency feature reference value range determined in real time. S53, the real-time deviation direction mode vector Perform pattern matching with the condition part of each rule in the mapping rule base, and calculate the matching degree; S54. The abnormal cause corresponding to the rule with the highest matching degree is used as the inferred cause of the real-time energy efficiency anomaly, and the energy efficiency anomaly diagnosis result is output; when the matching degree of the rule with the highest matching degree is lower than the preset matching degree threshold, the real-time deviation direction mode vector is marked as a new mode to update the mapping rule library.

7. A system for implementing the energy consumption and energy efficiency analysis method according to any one of claims 1-6, characterized in that, It includes a processor and a memory, the memory being used to store a computer program, which, when executed by the processor, implements the energy consumption and energy efficiency analysis method as described in any one of claims 1-6.

Images