Wind disc fault monitoring method and system

Abstract

The invention belongs to the technical field of data processing, and particularly relates to a wind disc fault monitoring method and system, and the method comprises the steps: collecting wind disc operation data and environment parameters in real time, constructing an environment response sensitivity tensor, calculating and eliminating environment-induced supply and return water temperature difference deviation, and obtaining a de-environmental operation feature vector. And in combination with the static hydraulic topology and the dynamic behavior difference, generating a dynamic weight between the wind discs, constructing a self-evolution covariance matrix, and realizing dynamic optimization calculation of the Mahalanobis distance. And fault type discrimination is carried out based on the spatial topological relation of the abnormal wind disc, monomer blockage and system-level blockage are accurately distinguished, and intelligent early warning is realized. The method can effectively inhibit the environment disturbance false alarm, and improves the fault positioning accuracy.

Description

TECHNICAL FIELD

[0002] The present application relates to the technical field of data processing. More particularly, the present application relates to a kind of wind disc fault monitoring method and system. BACKGROUND

[0004] In modern intelligent building, fan coil as central air conditioning terminal equipment, undertakes the key task of adjusting indoor thermal comfort. Its running state directly affects user feeling, energy efficiency and system life. With the development of building intelligence, data-driven fault monitoring technology gradually replaces traditional manual inspection mode, and becomes the core means to improve operation efficiency.

[0005] Currently, for common faults of fan coil system (such as return air filter blockage, water filter blockage, etc.), the industry widely uses multivariate statistical analysis method for anomaly detection. Among them, the multivariate anomaly detection algorithm based on Mahalanobis distance is widely used in equipment health assessment and fault warning scene because it can effectively measure the deviation degree between equipment state and normal distribution of group. This method constructs the covariance matrix of the running data between fan coils, calculates the Mahalanobis distance between the state vector of a single device and the mean value of the group, and then identifies abnormal individuals.

[0006] In fan coil system anomaly detection, although the traditional Mahalanobis distance method can effectively identify the trend of equipment deviating from group behavior, when outdoor temperature suddenly changes or solar radiation increases, all fan coils will cause the supply and return water temperature difference to increase due to synchronous load increase, and the Mahalanobis distance will misjudge this kind of normal response as system-level blockage. At the same time, the static covariance model of Mahalanobis distance cannot adapt to the running drift caused by seasonal change or use mode change, resulting in poor robustness of diagnosis result, which is easy to cause false detection or missed detection of air conditioning fan coil system fault. SUMMARY

[0008] In order to solve the technical problems that the above-mentioned Mahalanobis distance method cannot distinguish between group normal response caused by environmental disturbance and real fault, and is difficult to adapt to long-term evolution trend of system, and is easy to produce false detection or missed detection, the present application provides solutions in the following aspects.

[0009] In a first aspect, the present application provides a fan coil fault monitoring method, comprising:

[0010] Real-time collection of air conditioner fan disc system operation data and environmental parameters, the operation data including water supply temperature and return water temperature; constructing an environmental response sensitivity tensor of the fan disc according to the dynamic response relationship between the fan disc operation data and the environmental parameters; determining an environmental-induced fan disc supply-return water temperature difference offset according to the deviation of the current environmental parameters from the historical baseline environmental parameters and the environmental response sensitivity tensor of the fan disc; correcting the supply-return water temperature difference of the fan disc according to the offset to obtain the supply-return water temperature difference after environmentalization, and constructing an operation feature vector of the fan disc in combination with the supply-return water temperature difference change rate; calculating the dynamic weight between the fan discs based on the static hydraulic topological relationship between the fan discs and the operation feature vector difference; constructing a self-evolution covariance matrix by using the dynamic weight, the operation feature vector and the historical covariance matrix; calculating the Mahalanobis distance based on the self-evolution covariance matrix and the operation feature vector of the fan disc, screening the abnormal fan disc, and discriminating the fan disc fault type and outputting the early warning according to the static hydraulic topological relationship between the abnormal fan discs.

[0011] Preferably, the environmental parameters include outdoor temperature, sunshine intensity and outdoor humidity.

[0012] Preferably, when constructing the environmental response sensitivity tensor of the fan disc, historical operation data and environmental parameters are used to realize by a linear regression model, the input of the linear regression model being an environmental variable vector and the output being the supply-return water temperature difference of the fan disc as a response variable.

[0013] Preferably, the determination of the environmental-induced fan disc supply-return water temperature difference offset includes: taking the difference between the current time environmental variable vector and the historical environmental variable vector mean as the current time environmental deviation; and taking the product of the environmental response sensitivity tensor of the fan disc and the current time environmental deviation as the current time environmental-induced fan disc supply-return water temperature difference offset.

[0014] Preferably, the operation feature vector is , wherein, is the operation feature vector of the i-th fan disc at the current time, is the supply-return water temperature difference of the i-th fan disc at the current time, is the supply-return water temperature difference of the i-th fan disc after environmentalization at the current time, , is the supply-return water temperature difference of the i-th fan disc at the current time, is the offset of the supply-return water temperature difference of the i-th fan disc induced by the environment at the current time, is the supply-return water temperature difference change rate of the i-th fan disc at the current time, is the supply-return water temperature difference of the i-th fan disc at the previous sampling time, is the supply-return water temperature difference of the i-th fan disc at the previous sampling time, is the supply-return water temperature difference of the i-th fan disc at the previous sampling time, , is the supply-return water temperature difference of the i-th fan disc at the previous sampling time, is the supply-return water temperature difference of the i-th fan disc at the previous sampling time, The temperature difference between the supply and return water of the typhoon control unit. This indicates a fixed sampling period.

[0015] Preferably, the dynamic weights satisfy the expression: ;in, Indicates the current moment. Indicates the current time. Typhoon Pan and the first Dynamic weights among typhoon disks; This indicates the static hydraulic connection weight. The static hydraulic connection weight between fan coil units on the same branch is set to 1, the static hydraulic connection weight between fan coil units that are on different branches but in the same unit is set to 0.3, and the static hydraulic connection weight between fan coil units in different units is set to 0. Indicates the first Typhoon Pan and the first The typhoon's dynamic behavior and distance over the past 24 hours; This is the attenuation coefficient.

[0016] Preferably, the dynamic behavior distance satisfies the expression: ;in, Indicates the first Typhoon Pan and the first The typhoon's dynamic behavior and distance over the past 24 hours; Indicates time Time The operational characteristic vector of the typhoon disk; Indicates time Time The operational characteristic vector of the typhoon disk; Represents the L2 norm; Indicates the length of the time window.

[0017] Preferably, the self-evolutionary covariance matrix satisfies the expression: ;in, Indicates the current moment. This represents the self-evolutionary covariance matrix at the current moment; Indicates the previous sampling time The self-evolutionary covariance matrix; Indicates the current time. Typhoon Pan and the first Dynamic weights among typhoon disks; Indicates the current time. The operational characteristic vector of the typhoon disk; Indicates the current time. The operational characteristic vector of the typhoon disk; This represents the mean of the operating characteristic vectors of all wind turbines at the current moment; This represents the matrix transpose operation; Indicates the learning rate; This indicates a fixed sampling period.

[0018] Preferably, the screening of abnormal fan coil units involves identifying the fault type and outputting an early warning based on the static hydraulic topology relationship between the abnormal fan coil units. This includes: determining that a fan coil unit has a potential abnormality if its Mahalanobis distance is greater than a preset threshold; determining that the main filter of the fan coil unit is clogged if only a single fan coil unit has a potential abnormality while the other fan coil units in the same branch are normal; determining that the filter of the branch is clogged if three or more fan coil units in the same branch have potential abnormalities simultaneously; and determining that the main filter of the unit is clogged if fan coil units in multiple branches have potential abnormalities simultaneously. Structured early warning information is generated based on the identification results and pushed to the operation and maintenance platform.

[0019] Secondly, the present invention provides a fan coil unit fault monitoring system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned fan coil unit fault monitoring method is implemented.

[0020] By adopting the above technical solution, a computer program for the above-mentioned fan coil unit fault monitoring method is generated and stored in a memory so that it can be loaded and executed by a processor. A terminal device can then be made based on the memory and processor for convenient use.

[0021] The beneficial effects of this invention are as follows: By constructing an environmental response sensitivity tensor and calculating environmental induced offsets, this invention effectively eliminates the interference of environmental factors on the operation data of the air handling units, avoiding misjudgments caused by changes in outdoor temperature, humidity, and sunlight, and significantly improving the environmental adaptability of anomaly detection; by integrating static hydraulic topology relationships and dynamic behavioral distances to calculate dynamic weights, and constructing a self-evolving covariance matrix based on this, this invention enables the model to perceive the physical relationships between air handling units and adapt to the long-term evolution trend of the system, solving the problem that traditional static models are unable to cope with seasonal changes and changes in usage patterns; by combining the self-evolving covariance matrix to calculate Mahalanobis distance, this invention can more accurately measure the deviation of the air handling unit's operating status from the normal distribution of the group, greatly improving the accuracy of anomaly identification; finally, this invention uses static hydraulic topology relationships to identify the fault type of abnormal air handling units, accurately locating faults such as blockage of the main filter, branch filter, or unit main filter, providing clear guidance for operation and maintenance, reducing manual troubleshooting costs, and improving the intelligent operation and maintenance level and energy utilization efficiency of the air conditioning air handling unit system. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a fan coil unit failure monitoring method according to the present invention. Detailed Implementation

[0024] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of the present application.

[0025] The specific embodiments of the present application will be described in detail below with reference to the drawings.

[0026] The embodiments of the present application disclose a wind disc fault monitoring method, referring to Figure 1 , comprising steps S1-S8:

[0027] S1, collecting the running data and environmental parameters of the air conditioner wind disc system in real time.

[0028] It should be noted that the running state of the air conditioner wind disc system is affected by multiple factors and will exhibit different characteristics under different environmental conditions and running periods. For example, the rise and fall of outdoor temperature will cause the supply and return water temperature difference of the wind disc to change, and the adjustment of the fan speed gear will also affect the heat exchange efficiency of the wind disc. Therefore, the present application collects the running data and environmental parameters of the wind disc system in real time to provide basic data support for subsequent anomaly detection.

[0029] Specifically, from the start of the wind disc system running, the intelligent controller and sensor network deployed on each wind disc are used to continuously collect running data, and environmental parameters are obtained through a roof weather station or a city weather interface. The collection frequency is set to 5 minutes once, and the implementer can adjust the collection frequency according to the actual situation.

[0030] The running data includes supply water temperature, return water temperature, etc. The supply water temperature is the water temperature at the inlet of the wind disc, and the return water temperature is the water temperature at the outlet of the wind disc. The environmental parameters include outdoor temperature, solar radiation intensity, outdoor humidity, etc.

[0031] Further, according to the collected supply water temperature and return water temperature, the supply and return water temperature difference can be calculated, wherein, represents the supply water temperature, represents the return water temperature, and the supply and return water temperature difference can reflect the heat exchange intensity of the wind disc.

[0032] Further, in order to eliminate the influence of sensor noise, communication jitter and transient interference on subsequent data analysis, the present application uses Kalman filter algorithm to carry out denoising processing on the collected running data and environmental parameters.

[0033] In other embodiments, the implementer can select other denoising methods according to the quality of the field data, such as wavelet transform, sliding average filtering, etc.

[0034] S2. Construct the environmental response sensitivity tensor of the fan coil unit based on the dynamic response relationship between the fan coil unit's operating data and environmental parameters.

[0035] It should be noted that different fan coil units respond differently to changes in environmental parameters. For example, fan coil units in south-facing rooms are more sensitive to changes in sunlight intensity, and in high humidity environments, the fan may increase airflow to dehumidify, thus affecting the operating status of the fan coil unit. Therefore, this invention quantifies the response characteristics of fan coil units to multi-dimensional environmental parameters by constructing an environmental response sensitivity tensor.

[0036] Specifically, using historical operating data and environmental parameters from the past 7 days, an environmental response sensitivity tensor is constructed for each wind turbine:

[0037] Construct the environment variable vector for each moment using all environmental parameters: ,in, For a moment outdoor temperature at that time For a moment Solar radiation intensity at that time For a moment The outdoor humidity at any given time. The supply and return water temperature difference of each fan coil unit at each moment is used as the response variable: ,in, For the first Typhoon Pan at all times The temperature difference between the supply and return water.

[0038] For any given fan coil unit, a linear regression model is established based on the environmental variable vector at each moment and the response variable of that fan coil unit:

[0039]

[0040] in, Indicates the first The environmental response sensitivity tensor of the typhoon disk. , Each element in the table represents the strength of the influence of a certain environmental parameter on the response variable, for example... Indicates the outdoor temperature for the first The impact of the temperature difference between the supply and return water of the typhoon control unit. Indicates the solar radiation intensity for the first The impact of the temperature difference between the supply and return water in the typhoon control system; For bias terms; A vector of environment variables; For the first The response variable vector of the typhoon disk.

[0041] It should be noted that the environmental response sensitivity tensor can accurately characterize the response characteristics of the wind turbine to multi-dimensional environmental inputs, laying the foundation for subsequent removal of environmental interference.

[0042] S3. Based on the deviation between the current environmental parameters and the historical baseline environmental parameters, as well as the environmental response sensitivity tensor of the fan coil unit, determine the offset of the environmentally induced supply and return water temperature difference of the fan coil unit.

[0043] It should be noted that changes in environmental parameters can cause deviations in the fan coil unit's operating data, which may be misjudged as abnormal. For example, an increase in outdoor temperature will increase the indoor cooling load, naturally leading to a larger temperature difference between the fan coil unit's supply and return water. Therefore, this invention calculates the environmentally induced deviation in the supply and return water temperature difference to isolate the influence of environmental factors on the supply and return water temperature difference.

[0044] Specifically, based on the deviation between current environmental parameters and historical baseline environmental parameters, and combined with the environmental response sensitivity tensor constructed in step S2, the offset of the environmentally induced supply and return water temperature difference of the fan coil unit is calculated, including:

[0045] Calculate the environmental deviation at the current moment based on the difference between the current environmental variable vector and the historical mean environmental variable vector:

[0046]

[0047] in, Indicates the current time; Indicates the environmental deviation at the current moment; This represents the vector of environment variables at the current moment. This represents the mean of the environmental variable vector over the past 7 days.

[0048] Furthermore, based on the environmental response sensitivity tensor and the environmental deviation at the current moment, the offset of the response variable is predicted:

[0049]

[0050] in, Indicates the current time; Indicates the environmentally induced first moment at the current time. The offset of the temperature difference between the supply and return water of the typhoon control plate; Indicates the first The environmental response sensitivity tensor of the typhoon disk; This indicates the environmental deviation at the current moment.

[0051] It should be noted that by calculating the offset of the supply and return water temperature difference of the fan coil unit induced by the environment, this invention can accurately distinguish between changes in operating data caused by environmental factors and changes in operating data caused by anomalies, thereby improving the accuracy of subsequent anomaly detection.

[0052] S4, according to the wind disc supply and return water temperature difference induced by the environment, the wind disc supply and return water temperature difference is corrected, the wind disc supply and return water temperature difference after environmentalization is obtained, and the wind disc operation characteristic vector is constructed according to the wind disc supply and return water temperature difference after environmentalization and the supply and return water temperature difference change rate.

[0053] It should be noted that after obtaining the environmental induction wind disc supply and return water temperature difference offset, it needs to be removed from the actual measured supply and return water temperature difference to obtain the“net heat transfer behavior”characteristics not affected by environmental factors, and provide pure input data for subsequent anomaly detection.

[0054] Specifically, according to the supply and return water temperature difference of each wind disc at the current moment and the environmental induction of the first Wind disc supply and return water temperature difference offset, the current moment environmentalization after the supply and return water temperature difference is determined:

[0055]

[0056] Wherein, Indicates the current moment; Indicates the current moment environmentalization after the supply and return water temperature difference of the first Wind disc; Indicates the current moment environmental induction of the first Wind disc supply and return water temperature difference offset; Indicates the current moment supply and return water temperature difference of the first Wind disc.

[0057] Further, the supply and return water temperature difference change rate is calculated:

[0058]

[0059] Wherein, Indicates the current moment, Indicates the current moment supply and return water temperature difference change rate of the first Wind disc, used to represent the dynamic trend of wind disc heat transfer process; Indicates the current moment supply and return water temperature difference of the first Wind disc; Indicates the previous sampling moment Supply and return water temperature difference of the first Wind disc; Indicates the fixed sampling period, in the embodiment Minutes, consistent with the system data acquisition frequency, to ensure the alignment of time series data and calculation stability.

[0060] Further, the current moment environmentalization after the supply and return water temperature difference of each wind disc and the current moment supply and return water temperature difference change rate of each wind disc are constructed into the operation characteristic vector of each wind disc at the current moment: wherein, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point.

[0061] It should be noted that the running feature vector of each air-cooled heat exchanger can reflect the real heat exchange behavior of the air-cooled heat exchanger, and the interference of environmental factors is excluded, thereby providing a reliable basis for subsequent group-based anomaly detection.

[0062] S5, calculating the dynamic weight between the air-cooled heat exchangers based on the static hydraulic topological relationship and the difference in the running feature vectors between the air-cooled heat exchangers.

[0063] It should be noted that there is a hydraulic coupling characteristic between each air-cooled heat exchanger in the air-cooled heat exchanger system, and the functional similarity between the air-cooled heat exchangers changes dynamically with the operation mode. For example, during the night low load period, the air-cooled heat exchangers in the east and west wings may exhibit stronger correlation. Therefore, the dynamic weight between the air-cooled heat exchangers is constructed according to the static hydraulic topological relationship and the dynamic behavior distance between the air-cooled heat exchangers, so as to quantify the correlation degree between different air-cooled heat exchangers.

[0064] Specifically, the dynamic behavior distance between different air-cooled heat exchangers is determined according to the difference between the running feature vectors of different air-cooled heat exchangers.

[0065]

[0066] wherein, represents the dynamic behavior distance between the i-th air-cooled heat exchanger and the j-th air-cooled heat exchanger within the past 24 hours; represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the running feature vector of the i-th air-cooled heat exchanger at the current time point, represents the L2 norm, which is used to calculate the spatial distance between two feature vectors; represents the length of the time window, (corresponding to 24 hours of data, with an interval of 5 minutes, i.e. 12 5-minute sampling points are contained in each hour), which is used to limit the time range for calculating the dynamic behavior distance.

[0067] Further, the dynamic weight between different air-cooled heat exchangers is determined according to the dynamic behavior distance between different air-cooled heat exchangers and the static hydraulic connection weight between the air-cooled heat exchangers:

[0068]

[0069] wherein, denotes the current time, denotes the dynamic weight between the current time, typhoon disk and the first typhoon disk; the dynamic weight between the current time, denotes the static hydraulic connection weight, in the embodiment, the static hydraulic connection weight between the wind disks of the same branch is set to 1, the static hydraulic connection weight between the wind disks of the cross branch but the same unit is set to 0.3, and the static hydraulic connection weight between the wind disks of different units is set to 0; denotes the dynamic behavior distance between the current time, typhoon disk and the first typhoon disk in the past 24 hours; the dynamic weight between the current time, is a decay coefficient, used for controlling the influence strength of the dynamic behavior distance on the dynamic weight, in the embodiment , in other embodiments, the implementer can set according to the actual implementation situation.

[0070] It should be noted that the dynamic weight fuses the static hydraulic topological relationship and the behavior similarity of the wind disk, and can more accurately reflect the correlation degree between the wind disks.

[0071] S6, according to the dynamic weight between the wind disks and the operation characteristic vector of the wind disk, a self-evolution covariance matrix is constructed in combination with a historical covariance matrix.

[0072] It should be noted that the operation behavior of the wind disk system has significant time-varying and non-stationary characteristics, and a static statistical model cannot adapt to the long-term evolution trend. Therefore, the sliding window weighting and online learning mechanism are adopted to construct the self-evolution covariance matrix, so that the model can adapt to the dynamic change of the system.

[0073] Specifically, the update formula of the self-evolution covariance matrix is:

[0074]

[0075] wherein, denotes the current time, denotes the self-evolution covariance matrix at the current time, used for describing the joint distribution characteristics of the de-environmentalized characteristic vectors in the group; denotes the self-evolution covariance matrix at the previous sampling time, used for retaining the statistical characteristics of the historical operation mode; denotes the dynamic weight between the current time, typhoon disk and the first typhoon disk, reflecting the correlation strength of the two in the structural topology and the behavior similarity; denotes the dynamic weight between the current time, The operational characteristic vector of the typhoon disk; Indicates the current time. The operational characteristic vector of the typhoon disk; This represents the mean of the operating characteristic vectors of all wind turbines at the current moment, reflecting the average operating state of the group as a whole; This represents the matrix transpose operation; The learning rate, ranging from [0,1], controls the fusion ratio of historical information and current data. In this embodiment... In other embodiments, implementers can set the learning rate according to the actual implementation situation. When the learning rate The larger the value, the more adaptable the model is to the latest behavior; This indicates a fixed sampling period, as shown in this embodiment. Minutes, consistent with the system's data acquisition frequency.

[0076] It should be noted that the present invention is achieved through... Preserving stable patterns in the historical covariance matrix avoids interference from short-term fluctuations on the model, while also leveraging... By incorporating weighted statistical information at the current moment, the latest operating trends can be tracked, achieving a fusion of historical and real-time data. Simultaneously, this invention utilizes the dynamic weights among the fan coil units. Deviation items between fan coil units By performing weighted summation, the contribution of physically related and behaviorally similar wind turbine devices to the covariance matrix is ​​enhanced, improving the model's ability to perceive topological structures. This allows the self-evolving covariance matrix to retain the stability of historical data while incorporating the latest behavioral patterns, effectively resisting short-term noise interference, adapting to long-term changes such as seasonal changes and usage pattern changes, and ensuring that the model maintains high detection accuracy at different times. This provides a reliable distribution model foundation for subsequent Mahalanobis distance calculations.

[0077] It is important to note that in the actual deployment of air conditioning fan coil systems, there must be at least one fan coil combination with a non-zero static weight (i.e., at least one pair of fan coils must satisfy the topological relationship of being on the same branch or in the same unit across branches). This deployment characteristic ensures that the denominator in the formula... The value is not zero, thus ensuring the validity of the calculation of the self-evolutionary covariance matrix.

[0078] S7. Based on the self-evolutionary covariance matrix and the operating feature vector of the wind turbine, calculate the Mahalanobis distance and filter out abnormal wind turbines.

[0079] It should be noted that Mahalanobis distance can measure the degree of deviation of a single sample from the sample set. Therefore, this invention uses Mahalanobis distance to identify wind turbine anomalies.

[0080] Specifically, the Mahalanobis distance of each wind turbine is determined based on its operational characteristic vector and self-evolutionary covariance matrix.

[0081]

[0082] in, Indicates the current moment. Indicates the current time. The Mahalanobis distance of the typhoon disk reflects the current moment. The degree to which the typhoon's operating status deviates from the normal operating status of the group is used to measure the... Is the typhoon warning system abnormal? Indicates the current time. The operational characteristic vector of the typhoon disk; This represents the mean of the operating characteristic vectors of all wind turbines at the current moment; This represents the self-evolutionary covariance matrix at the current moment; This represents the matrix transpose operation.

[0083] when The larger the value, the more significant the first... The greater the deviation between the operating status of a typhoon manipulator and the normal operating status of the group, the higher the probability that the manipulator is abnormal; conversely, when The smaller the value, the closer the operation of the wind turbine is to the average level of the group, and the more normal its operation. This invention calculates the deviation between the individual wind turbine's operational characteristic vector and the group's mean vector, and then weights this deviation using the inverse of the self-evolutionary covariance matrix, thus obtaining a quantitative indicator that reflects the difference between the individual and the overall group's operational patterns. (Self-evolutionary covariance matrix) It incorporates the topological relationships and dynamic behavior characteristics of the wind turbine system. Its inverse matrix can reasonably scale the differences in features of different dimensions, avoiding the interference of features of different dimensions on the distance calculation. This makes the obtained Mahalanobis distance consider both the absolute differences between individual wind turbines and the group, as well as the distribution pattern of features within the group, and can more accurately identify the abnormal operating state of the wind turbine.

[0084] Furthermore, in response to Mahalanobis distance Greater than the preset threshold The system determines that the fan coil unit has a potential malfunction and proceeds to the fault type identification stage. In this embodiment, the threshold... The empirical value is 3.2. In other embodiments, implementers can set the threshold according to the actual implementation situation. .

[0085] S8. Based on the static hydraulic topology relationship between abnormal fan coil units, determine the type of fan coil unit failure and output early warning.

[0086] It should be noted that for the wind disc marked as potential abnormality, the application combines the topological position of the wind disc in the system and the abnormal conditions of other wind discs to determine the fault type, so as to determine the specific fault position and type and provide accurate guidance for operation and maintenance.

[0087] Specifically, in response to only one wind disc existing potential abnormality, the remaining wind discs of the same branch are normal, it is determined that the body filter of the wind disc is blocked; in response to three or more wind discs in the same branch synchronously existing potential abnormality, it is determined that the branch filter is blocked; in response to the wind discs in multiple branches simultaneously existing potential abnormality, it is determined that the unit total filter is blocked.

[0088] Further, structured early warning information is generated according to the determination result and is pushed to an operation and maintenance platform.

[0089] The embodiment of the application further discloses a wind disc fault monitoring system, comprising a processor and a memory, and the memory stores computer program instructions, when the computer program instructions are executed by the processor, a wind disc fault monitoring method according to the application is realized.

[0090] The above system further comprises a communication bus and a communication interface and other components familiar to those skilled in the art, the settings and functions of which are known in the art, and thus will not be described here.

[0091] In the description of the present specification, the meaning of "a plurality of", "several" is at least two, for example, two, three or more, etc., unless otherwise explicitly specifically limited.

[0092] Although the present specification has shown and described several embodiments of the present application, it will be apparent to those skilled in the art that such embodiments are provided only in an exemplary manner. Those skilled in the art will think of many changes, changes and alternative ways without departing from the idea and spirit of the present application. It should be understood that in the process of practicing the present application, each alternative to the embodiments of the present application described herein can be adopted.

Claims

1. A method for monitoring fan coil unit failures, characterized in that, include: Real-time acquisition of operating data and environmental parameters of the air conditioning fan coil system, including water supply temperature and return water temperature; An environmental response sensitivity tensor for the fan is constructed based on the dynamic response relationship between the fan operating data and environmental parameters. The environmentally induced offset of the fan supply and return water temperature difference is determined based on the deviation between the current environmental parameters and historical baseline environmental parameters, as well as the environmental response sensitivity tensor of the fan. The fan supply and return water temperature difference is corrected based on the offset to obtain the de-environmentalized fan supply and return water temperature difference. The operating feature vector of the fan is constructed by combining the rate of change of the fan supply and return water temperature difference. The dynamic weights between the wind turbines are calculated based on the static hydraulic topology relationship and the difference in their operating feature vectors. An evolutionary covariance matrix is ​​constructed using the dynamic weights, operating feature vectors, and historical covariance matrices. Based on the evolutionary covariance matrix and the operating feature vectors of the wind turbines, Mahalanobis distance is calculated to screen abnormal wind turbines. Based on the static hydraulic topology relationship between the abnormal wind turbines, the fault type of the wind turbines is identified and an early warning is output.

2. The method for monitoring fan coil unit failure according to claim 1, characterized in that, The environmental parameters include outdoor temperature, solar radiation intensity, and outdoor humidity.

3. The method for monitoring fan coil unit failure according to claim 1, characterized in that, When constructing the environmental response sensitivity tensor of the fan coil unit, historical operating data and environmental parameters are used to achieve this through a linear regression model. The input of the linear regression model is an environmental variable vector, and the output is the supply and return water temperature difference of the fan coil unit as the response variable.

4. The method for monitoring fan coil unit failure according to claim 1, characterized in that, The determination of the environmentally induced supply and return water temperature difference offset includes: The difference between the current environmental variable vector and the mean of the historical environmental variable vector is taken as the environmental deviation at the current moment; the product of the environmental response sensitivity tensor of the fan coil unit and the environmental deviation at the current moment is taken as the offset of the supply and return water temperature difference of the fan coil unit induced by the environment at the current moment.

5. The method for monitoring fan coil unit failure according to claim 1, characterized in that, The running feature vector is ,in, Indicates the current time. The operational characteristic vector of the typhoon disk. Indicates the current time. The temperature difference between the supply and return water after the typhoon shield is de-environmentally treated: , Indicates the current time. The temperature difference between the supply and return water of the typhoon control unit. Indicates the environmentally induced first moment at the current time. The offset of the temperature difference between the supply and return water of the typhoon control plate; Indicates the current time. The rate of change of supply and return water temperature difference in the typhoon control unit: , Indicates the previous sampling time No. The temperature difference between the supply and return water of the typhoon control unit. This indicates a fixed sampling period.

6. The method for monitoring fan coil unit failure according to claim 1, characterized in that, Obtain dynamic weights that satisfy the expression: ； in, Indicates the current moment. Indicates the current time. Typhoon Pan and the first Dynamic weights among typhoon disks; This indicates the static hydraulic connection weight. The static hydraulic connection weight between fan coil units on the same branch is set to 1, the static hydraulic connection weight between fan coil units that are on different branches but in the same unit is set to 0.3, and the static hydraulic connection weight between fan coil units in different units is set to 0. Indicates the first Typhoon Pan and the first The dynamic behavior and distance of the typhoon over the past 24 hours; This is the attenuation coefficient.

7. The method for monitoring fan coil unit failure according to claim 6, characterized in that, The dynamic behavior distance satisfies the expression: ； in, Indicates the first Typhoon Pan and the first The dynamic behavior and distance of the typhoon over the past 24 hours; Indicates time Time The operational characteristic vector of the typhoon disk; Indicates time Time The operational characteristic vector of the typhoon disk; Represents the L2 norm; Indicates the length of the time window.

8. The method for monitoring fan coil unit failure according to claim 1, characterized in that, The self-evolutionary covariance matrix satisfies the expression: ； in, Indicates the current moment. This represents the self-evolutionary covariance matrix at the current moment; Indicates the previous sampling time The self-evolutionary covariance matrix; Indicates the current time. Typhoon Pan and the first Dynamic weights among typhoon disks; Indicates the current time. The operational characteristic vector of the typhoon disk; Indicates the current time. The operational characteristic vector of the typhoon disk; This represents the mean of the operating characteristic vectors of all wind turbines at the current moment; This represents the matrix transpose operation; Indicates the learning rate; This indicates a fixed sampling period.

9. The method for monitoring fan coil unit failure according to claim 1, characterized in that, The process of screening abnormal fan coil units involves identifying the type of fan coil unit failure and outputting an early warning based on the static hydraulic topology relationship between the abnormal fan coil units, including: If the Mahalanobis distance of the fan coil unit is greater than a preset threshold, it is determined that the fan coil unit has a potential anomaly. If only a single fan coil unit has a potential anomaly while the rest of the fan coil units in the same branch are normal, it is determined that the filter of that fan coil unit is blocked. If three or more fan coil units in the same branch have potential anomalies simultaneously, it is determined that the filter of that branch is blocked. If fan coil units in multiple branches have potential anomalies simultaneously, it is determined that the main filter of the unit is blocked. Based on the judgment results, structured early warning information is generated and pushed to the operation and maintenance platform.

10. A fan coil unit fault monitoring system, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a fan coil unit fault monitoring method according to any one of claims 1-9.

Images