Intelligent control method for energy-saving water supply and drainage system of building

Abstract

The invention discloses an intelligent control method for a building energy-saving water supply and drainage system, and the method comprises the steps: constructing a multivariate state feature vector of each drainage pump, and outputting a plurality of drainage pump groups; based on each draining pump group, calculating a state extreme value of each draining pump in each historical time window to form a state fluctuation sequence and a state deviation sequence of other draining pumps in the draining pump group; if the number of historical time windows reaching a deviation threshold value in a certain state deviation sequence is greater than a preset number, determining the state deviation sequence as a suspected abnormal drainage pump; on the basis of each suspected abnormal drainage pump, a multivariate deviation set and a corresponding abnormal probability value of the suspected abnormal drainage pump are obtained every other period, and an abnormal probability value sequence of each suspected abnormal drainage pump is generated; and generating a hit rate of the abnormal probability value sequence, and if the hit rate is greater than a hit threshold, determining that the suspected abnormal draining pump is diagnosed to be abnormal. Therefore, the reliability and the accuracy of abnormal detection of the draining pump are improved, and the detection efficiency is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to a control method for a building energy-saving water supply and drainage system. Background Art

[0002] The building water supply and drainage system is used to remove sewage from residential buildings, public buildings and production buildings. The drainage system inside the building is generally composed of water receivers for sanitary appliances or production equipment, drainage pipes, clearing facilities, ventilation pipes, sewage and wastewater lifting equipment and local treatment structures. In the building water supply and drainage system, the drainage pump plays an irreplaceable role. Reasonable control of the drainage pump in the building water supply and drainage system can provide guarantee for the stable operation of the building water supply and drainage system. When the drainage pump is abnormal, it will show abnormal vibration.

[0003] At present, the common method for detecting abnormal conditions of drainage pumps and controlling drainage pumps is to determine whether the drainage pump is working abnormally by detecting the change in vibration data of each drainage pump and the vibration data of a normal drainage pump. Since the drainage pump has vibrations when it is working, the abnormal vibration is more obvious, that is, when the abnormality is more serious, the abnormality of the drainage pump can be determined by only detecting the vibration data of the drainage pump; however, when the drainage pump is slightly abnormal, the abnormal vibration is not obvious, and it is difficult to distinguish the abnormality of the drainage pump by only comparing it with the vibration data of a normal drainage pump.

[0004] The Chinese invention patent with patent application number 202210818029.5 discloses an intelligent control method for a building energy-saving water supply and drainage system. Based on the multi-segment vibration sequence of each drainage pump, the common factor matrix and independent factor vector of each drainage pump are obtained, and a common scatter plot and an independent scatter plot are constructed to obtain the degree of abnormality of the drainage pump. By conducting an overall analysis of the working status and independent parameters of the drainage pumps working simultaneously, and combining the independent status of the drainage pumps for separate analysis, slight abnormalities in the drainage pump can be detected in time.

[0005] However, there may be certain limitations in relying solely on the vibration sequence of the drainage pump for abnormality judgment, because the vibration sequence may be affected by many factors, including water flow abnormalities caused by external factors at the drainage end and internal factors of the drainage pump itself. These factors may cause occasional vibration abnormalities, but the drainage pump itself is not abnormal, thus resulting in abnormal diagnosis of the drainage pump. Summary of the invention

[0006] The present application provides an intelligent control method for a building energy-saving water supply and drainage system, thereby improving the reliability and accuracy of drainage pump abnormality detection while enhancing detection efficiency.

[0007] The present application provides an intelligent control method for a building energy-saving water supply and drainage system, comprising:

[0008] S101, collecting vibration characteristic data and working characteristic parameters of all drainage pumps in each historical time window within a preset historical time period, and constructing a multivariate state characteristic vector of each drainage pump;

[0009] S102, inputting the multivariate state feature vectors of all drainage pumps into a preset clustering algorithm, and outputting K clusters, each cluster corresponding to a drainage pump group, and each drainage pump group including at least one drainage pump;

[0010] S103, based on each drainage pump group, calculating the state extreme value of each drainage pump in each historical time window to form a state fluctuation sequence of the drainage pump, and generating a state deviation sequence of the drainage pump from each other drainage pump in the drainage pump group;

[0011] S104, among all state deviation sequences of each drainage pump in the drainage pump group, if there is a state deviation sequence in which the number of historical time windows reaching the deviation threshold is greater than a preset number, the corresponding drainage pump is determined as a suspected abnormal drainage pump;

[0012] S105, based on each suspected abnormal drainage pump, obtaining its multivariate deviation set and its corresponding abnormal probability value every preset period, and generating an abnormal probability value sequence for each suspected abnormal drainage pump;

[0013] S106, based on each suspected abnormal drainage pump, generate a hit rate of the abnormal probability value sequence according to a preset abnormal probability threshold and abnormal probability value sequence. If the hit rate is greater than the hit threshold, the suspected abnormal drainage pump is determined to be abnormal.

[0014] Preferably, the vibration characteristic data includes vibration frequency, amplitude, and vibration duration, the working characteristic parameters include flow, pressure, rotation speed, and temperature, and the multivariate state characteristic vector of each drainage pump is constructed, including:

[0015] A1. Calculate the statistics of vibration characteristic data, including the mean, variance, and kurtosis values ​​corresponding to vibration frequency, amplitude, and vibration duration respectively;

[0016] A2. Calculate the statistics of the working characteristic parameters, including the mean, variance, and kurtosis values ​​corresponding to flow, pressure, speed, and temperature respectively;

[0017] A3. Combine the statistics of the vibration characteristic data and the statistics of the working characteristic parameters to generate a multivariate state characteristic vector of each drainage pump.

[0018] Preferably, the pre-set clustering algorithm is set to a K-Means clustering algorithm, and the center point of each cluster is determined as the label feature vector of the corresponding drainage pump group.

[0019] Qk], the state deviation sequence is expressed as [ΔQ1, ΔQ2, ..., ΔQk];

[0020] Qk is the state extreme value in the kth historical time window within the preset historical period, and ΔQk is the state deviation value of the drainage pump from other drainage pumps in the drainage pump group.

[0021] Preferably, the method for calculating the state extreme value is:

[0022] The vibration extreme value and working extreme value corresponding to the vibration characteristic data and working characteristic parameters of each drainage pump in each historical time window are obtained, and the vibration extreme value and the working extreme value are weightedly summed to obtain the state extreme value.

[0023] Preferably, obtaining the multivariate state deviation set and its corresponding abnormal probability value every preset period includes:

[0024] B1. Based on each suspected abnormal drainage pump, obtain the multivariate state feature vector in the previous cycle and the multivariate state feature vector set of all other drainage pumps in the drainage pump group at every preset cycle;

[0025] B2. Calculate the Euclidean distance value between the drainage pump and each multivariate state feature vector in the multivariate state feature vector set respectively to form a multivariate state deviation set of the drainage pump. Each element in the set corresponds to a Euclidean distance value. The ratio of the number of elements in the multivariate state deviation set that are greater than a preset distance threshold to the total number of elements is determined as the abnormal probability value.

[0026] Preferably, the hit rate of the abnormal probability value sequence is set to: the number of abnormal probability values ​​greater than the abnormal probability threshold / the total number of abnormal probability values ​​in the abnormal probability value sequence.

[0027] Preferably, the S106 includes:

[0028] When the hit rate of the abnormal probability value sequence is within the preset first interval, the suspected abnormal drainage pump is determined as the target drainage pump, and the suspected abnormal drainage pump at the corresponding point is re-inspected for abnormality using the preset abnormal retest mechanism to output its final diagnosis result; wherein the first interval is set to: [30%, 90%].

[0029] Preferably, the pre-set abnormal retest mechanism specifically includes:

[0030] S201, obtaining the Euclidean distance value of the multivariate state feature vector of the target drainage pump and other drainage pumps diagnosed as normal in the drainage pump group in which the target drainage pump is located within a recent preset period;

[0031] S202, inputting the Euclidean distance value into a pre-trained state abnormality prediction model, and outputting a diagnosis result of the target drainage pump, the diagnosis result including abnormality and normality;

[0032] Among them, each drainage pump group corresponds to a state abnormality prediction model. The state abnormality prediction model corresponding to the target drainage pump is determined according to the drainage pump group where the target drainage pump is located. The specific process of obtaining the state abnormality prediction model completed in advance is as follows:

[0033] In a drainage pump group, a large number of multivariate state feature vectors of all drainage pumps in historical periods are obtained, and the drainage pumps with long-term stable and normal operation in the drainage pump group are selected as reference objects. The Euclidean distance values ​​of the multivariate state feature vectors of each drainage pump and the corresponding reference object in all historical periods are obtained as training data sets, and each training data is labeled with a label indicating whether the state of the drainage pump is abnormal or normal; the pre-set neural network structure is trained using the training data set, and the model parameters are continuously optimized to obtain the final state abnormality prediction model.

[0034] Preferably, each of the drainage pump groups corresponds to a state abnormality prediction model, each state abnormality prediction model includes a plurality of differentiated prediction models, and the S202 further includes:

[0035] S301, based on the preset first interval, dividing into three sub-intervals [30%, 50%), [50%, 70%), and [70%, 90%), each sub-interval corresponding to a differentiated prediction model;

[0036] S302, obtaining the hit rate corresponding to the abnormal probability value sequence of the target drainage pump, and locating the corresponding differentiated prediction model according to the sub-interval where the hit rate is located;

[0037] S303: Input the Euclidean distance value corresponding to the target drainage pump into the corresponding differential prediction model, and output the diagnosis result of the target drainage pump.

[0038] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0039] By introducing multivariate data related to the working status of the drainage pump (such as working characteristic parameters) and refined analysis methods (such as cluster analysis, state extreme value and deviation sequence calculation), abnormal drainage pumps can be identified more accurately; by calculating the abnormal probability value sequence and hit rate, a more reliable basis for abnormal judgment is provided, reducing the possibility of misjudgment and missed judgment; cluster analysis based on multivariate state feature vectors divides the drainage pumps into similar groups, reducing the workload of comprehensive self-inspection, and only comparison, comparison and verification are required within the divided drainage pump groups, improving the efficiency and reliability of abnormality detection; the calculation of multivariate state feature vectors and state extreme values ​​provides comprehensive information on the changes in the operating status of the drainage pump, which helps to better understand its working status and abnormal causes. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flow chart of the intelligent control method of the building energy-saving water supply and drainage system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] To facilitate the understanding of the present invention, the present application will be described more comprehensively below with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, but the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to enable a more thorough and comprehensive understanding of the disclosed content of the present invention.

[0042] It should be noted that the terms “vertical”, “horizontal”, “up”, “down”, “left”, “right” and similar expressions used in this document are only for illustrative purposes and do not represent the only implementation method.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which the present invention belongs; the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention; the term "and / or" used herein includes any and all combinations of one or more related listed items.

[0044] Embodiment 1: Figure 1 It is a flow chart of an intelligent control method for a building energy-saving water supply and drainage system according to an embodiment of the present invention.

[0045] like Figure 1 As shown, a building energy-saving water supply and drainage system intelligent control method comprises the following steps:

[0046] S101, collect the vibration characteristic data and working characteristic parameters of all drainage pumps in each historical time window within a preset historical time period, the vibration characteristic data include vibration frequency, amplitude, vibration duration, the working characteristic parameters include flow rate, pressure, speed, temperature, pre-process the collected vibration characteristic data and working characteristic parameters, and construct a multivariate state characteristic vector for each drainage pump.

[0047] Specifically, the preset historical time period can be set to the past week, and the time window can be set to daily or to 6 hours, which can be adjusted according to actual conditions.

[0048] Among them, the collected vibration characteristic data and working characteristic parameters are preprocessed to construct the multivariate state characteristic vector of each drainage pump, including:

[0049] A1. Calculate the statistics of vibration characteristic data, including the mean, variance, and kurtosis (the difference between the low peak and the high peak) corresponding to the vibration frequency, amplitude, and vibration duration;

[0050] A2. Calculate the statistics of the working characteristic parameters, including the mean, variance, and kurtosis values ​​corresponding to flow, pressure, speed, and temperature respectively;

[0051] A3. Combine the statistics of the vibration characteristic data and the statistics of the working characteristic parameters to generate a multivariate state characteristic vector of each drainage pump.

[0052] S102, inputting the multivariate state feature vectors of all drainage pumps into a preset clustering algorithm, and outputting K clusters, each cluster corresponding to a drainage pump group, and each drainage pump group including at least one drainage pump.

[0053] Specifically, the pre-set clustering algorithm is set to the K-Means clustering algorithm, and the center point (central multivariate state feature vector) of each cluster is determined as the label feature vector of the corresponding drainage pump group.

[0054] Therefore, the drainage pump groups are divided based on the multivariate state feature vectors of all drainage pumps so that the drainage pumps in each group have similar comprehensive operating states, providing a reliable classification basis for subsequent self-inspection based on the data between the drainage pumps in the group. This can reduce the workload of comprehensive self-inspection and only requires comparison between the drainage pumps in each group.

[0055] S103, based on each drainage pump group, calculate the state extreme value of each drainage pump in each historical time window, [Qk], and generate a state deviation sequence [ΔQ1, ΔQ2, ..., Qk] of the drainage pump and each other drainage pump in the drainage pump group.

[0056] Qk is the state extreme value in the kth historical time window within the preset historical period, and ΔQk is the state deviation value of the drainage pump from other drainage pumps in the drainage pump group.

[0057] Specifically, the calculation method of the state extreme value is: obtain the vibration extreme value corresponding to the vibration characteristic data and working characteristic parameters of each drainage pump in each historical time window (set as: the difference between the maximum and minimum values ​​of all vibration characteristic data, take the average or weighted sum, used to reflect the influence of different vibration characteristic data), working extreme value (set as: the difference between the maximum and minimum values ​​of all working characteristic parameters, take the average or weighted sum, used to reflect the influence of different working characteristic parameters), and perform weighted summation of the vibration extreme value and the working extreme value to obtain the state extreme value. The weight values ​​of the vibration extreme value and the working extreme value are set according to the actual situation, which are used to indicate the importance of the vibration data and the working parameters to the drainage pump state assessment, and the sum of the two weight values ​​is 1.

[0058] S104: among all state deviation sequences of each drainage pump in the drainage pump group, if there is a state deviation sequence in which the number of historical time windows reaching the deviation threshold is greater than a preset number, the corresponding drainage pump is determined as a suspected abnormal drainage pump.

[0059] Among them, the deviation threshold is pre-set according to the actual situation and is used to measure the degree of deviation of the states between different drainage pumps in the same drainage pump group; the preset number is set according to the k value and can be set to half of the k value, indicating that when the number is greater than half, the drainage pump may be abnormal. This step performs a large-scale preliminary screening to screen out drainage pumps that may have abnormalities and determine them as suspected abnormal drainage pumps.

[0060] S105, based on each suspected abnormal drainage pump, obtain its multivariate deviation set and its corresponding abnormal probability value every preset period until the preset monitoring time is reached, and generate an abnormal probability value sequence for each suspected abnormal drainage pump. Wherein, the monitoring time is greater than the preset period, and the monitoring time is an integer multiple of the preset period, for example, the integer multiple can be set to 2.

[0061] Specifically, the multivariate state deviation set and its corresponding abnormal probability value are obtained every preset period, including:

[0062] B1. Based on each suspected abnormal drainage pump, obtain the multivariate state feature vector in the previous cycle and the multivariate state feature vector set of all other drainage pumps in the drainage pump group every preset cycle.

[0063] B2. Calculate the Euclidean distance value between the drainage pump and each multivariate state feature vector in the multivariate state feature vector set to form the multivariate state deviation set of the drainage pump. Each element in the set corresponds to a Euclidean distance value. The ratio of the number of elements in the multivariate state deviation set that are greater than the preset distance threshold to the total number of elements is determined as the abnormal probability value. The preset distance threshold is set according to the actual situation and is used to measure the degree of deviation of the multivariate state feature vectors of the two drainage pumps.

[0064] S106, based on each suspected abnormal drainage pump, generate the hit rate of the abnormal probability value sequence according to the preset abnormal probability threshold and abnormal probability value sequence. If the hit rate is greater than the hit threshold, the suspected abnormal drainage pump is determined to be abnormal, and the drainage control end is notified for maintenance.

[0065] The hit rate of the abnormal probability value sequence is set to: the number of abnormal probability values ​​greater than the abnormal probability threshold / the total number of abnormal probability values ​​in the abnormal probability value sequence. The hit threshold is set according to actual conditions, for example, to 90%.

[0066] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0067] By introducing multivariate data related to the working status of drainage pumps (such as working characteristic parameters) and refined analysis methods (such as cluster analysis, state extreme value and deviation sequence calculation), abnormal drainage pumps can be identified more accurately;

[0068] By calculating the abnormal probability value sequence and hit rate, a more reliable basis for abnormal judgment is provided, reducing the possibility of misjudgment and missed judgment;

[0069] Cluster analysis based on multivariate state feature vectors divides drainage pumps into similar groups, reducing the workload of comprehensive self-inspection. Only comparison and verification within the divided drainage pump groups are required, which improves the efficiency and reliability of anomaly detection.

[0070] The calculation of multivariate state eigenvectors and state extreme values ​​provides comprehensive information on the changes in the operating state of the drainage pump, which helps to better understand its working state and abnormal causes.

[0071] Embodiment 2: In embodiment 1, although the hit rate of the abnormal probability value sequence is used to determine whether the drainage pump is abnormal, due to the complexity and variability of the working environment of the drainage pump, there may be misjudgments or missed detections caused by the hit rate being close to the threshold. In particular, in some special cases, such as when the drainage pump is in a transitional state or is subjected to a short-term interference, its hit rate may fall near the determination threshold, making it difficult to directly determine whether it is abnormal.

[0072] Therefore, the embodiments of the present application are optimized to a certain extent based on the above embodiments.

[0073] In some embodiments, step S106 includes:

[0074] When the hit rate of the abnormal probability value sequence is within the preset first interval, the suspected abnormal drainage pump is determined as the target drainage pump, and the abnormal suspected abnormal drainage pump at the corresponding point is re-tested by using the preset abnormal retest mechanism, and the final diagnosis result is output. The first interval is set according to the actual situation, for example, it is set to: [30%, 90%].

[0075] In some embodiments, the pre-set abnormal retest mechanism specifically includes:

[0076] S201, obtaining the Euclidean distance value of the multivariate state feature vector between the target drainage pump and other drainage pumps diagnosed as normal in the drainage pump group in which the target drainage pump is located within a recent preset period.

[0077] S202, input the Euclidean distance value into a pre-trained state abnormality prediction model, and output a diagnosis result of the target drainage pump, where the diagnosis result includes abnormality and normality.

[0078] Among them, each drainage pump group corresponds to a state abnormality prediction model. The state abnormality prediction model corresponding to the target drainage pump is determined according to the drainage pump group where the target drainage pump is located. The specific process of obtaining the state abnormality prediction model completed in advance is as follows:

[0079] In a drainage pump group, a large number of multivariate state feature vectors of all drainage pumps in historical periods are obtained, and drainage pumps in the drainage pump group that have been operating stably and normally for a long time (determined by management personnel) are selected as reference objects. The Euclidean distance values ​​of the multivariate state feature vectors of each drainage pump and the corresponding reference object (model reference object in the same drainage pump group) in all historical periods are obtained as training data sets, and each training data is labeled (whether the drainage pump is actually normal or abnormal); the training data set is used to train the pre-set neural network structure, and the model parameters are continuously optimized to obtain the final state abnormality prediction model.

[0080] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0081] By setting the first interval, the hit rate of the abnormal probability value sequence is judged more carefully. When the hit rate falls within this interval, it is not directly judged whether the drainage pump is abnormal, but a further abnormal re-inspection is performed, which avoids misjudgment or missed judgment caused by the hit rate approaching the threshold, and improves the accuracy of abnormal judgment;

[0082] An abnormal retest mechanism was proposed. When the hit rate of the target drainage pump was within the first interval, the mechanism was used for retest. The retest process compared the Euclidean distance value of the multivariate state feature vector of the target drainage pump with that of other normal drainage pumps in the group, and input the state abnormality prediction model for judgment, thereby outputting a more reliable diagnostic result.

[0083] A state abnormality prediction model is trained for each drainage pump group. The model is trained based on the historical data of the drainage pumps in the group. The model can learn the normal operating status of the drainage pumps in the group and judge whether the target drainage pump is abnormal based on this, thereby improving the targetedness of abnormality identification.

[0084] Since the working environment of drainage pumps is complex and changeable, the adaptability of the method to different working environments is enhanced by training a state anomaly prediction model specific to the drainage pump group. This approach enables the method to better adapt to the operating characteristics of different drainage pump groups and improves the accuracy of anomaly identification.

[0085] Embodiment 3: In Embodiment 2, although the accuracy of abnormality identification is improved by re-inspecting the drainage pump by introducing the abnormal state prediction model, there is still a possibility of false detection due to the complexity and diversity of the working environment of the drainage pump and the diversity of abnormal characteristics. In particular, when the hit rate is close to the sub-interval boundary, a single abnormal state prediction model may not be able to accurately determine whether the drainage pump is abnormal.

[0086] Therefore, the embodiments of the present application are optimized to a certain extent based on the above embodiments.

[0087] In some embodiments, each drainage pump group corresponds to a state abnormality prediction model, each state abnormality prediction model includes a plurality of differentiated prediction models, and step S202 further includes:

[0088] S301, based on the preset first interval, divide it into three sub-intervals [30%, 50%), [50%, 70%), and [70%, 90%), each sub-interval corresponding to a differentiated prediction model.

[0089] S302, obtaining the hit rate corresponding to the abnormal probability value sequence of the target drainage pump, and locating the corresponding differentiated prediction model according to the sub-interval where the hit rate is located.

[0090] S303: input the Euclidean distance value corresponding to the target drainage pump into the corresponding differential prediction model, and output the diagnosis result of the target drainage pump.

[0091] In some embodiments, the training method of the differentiated prediction model corresponding to each sub-interval specifically includes:

[0092] S401, in the acquired training data set, obtaining the hit rate corresponding to the abnormal probability value sequence obtained by the drainage pump corresponding to each training data in the drainage pump group in which it is located.

[0093] S402, dividing each training data into corresponding sub-intervals according to the hit rate of each training data, and obtaining a sub-training set corresponding to each sub-interval.

[0094] S403, using the sub-training set to train and optimize the parameters of the preset neural network structure to obtain a differentiated prediction model corresponding to each sub-interval.

[0095] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0096] The abnormal state prediction model corresponding to each drainage pump group is refined into several differentiated prediction models. Each sub-interval corresponds to a differentiated prediction model, so that the model can process drainage pumps within different hit rate ranges more specifically, improving the adaptability and accuracy of the model.

[0097] The preset first interval [30%, 90%] is divided into three sub-intervals [30%, 50%), [50%, 70%), and [70%, 90%). Each sub-interval corresponds to a differentiated prediction model. Through this detailed division, the hit rate of the abnormal probability value sequence of the drainage pump can be more accurately located, so as to select a more appropriate differentiated prediction model for re-inspection.

[0098] Since drainage pumps in different hit rate ranges may have different abnormal characteristics, the drainage pumps in each sub-interval are re-inspected through a differentiated prediction model, which improves the accuracy and reliability of the re-inspection. This approach reduces false detections caused by insufficient model generalization ability and improves the accuracy of anomaly identification.

[0099] During the model training process, the training data is divided into corresponding sub-intervals according to their hit rates, and a sub-training set corresponding to each sub-interval is obtained. The sub-training set is used to train the preset neural network structure and optimize the parameters to obtain a differentiated prediction model corresponding to each sub-interval. This approach enhances the pertinence of model training and improves model performance.

[0100] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent control method for building energy-saving water supply and drainage system, characterized in that: include: S101, collecting vibration characteristic data and working characteristic parameters of all drainage pumps in each historical time window within a preset historical time period, and constructing a multivariate state characteristic vector of each drainage pump; S102, inputting the multivariate state feature vectors of all drainage pumps into a preset clustering algorithm, and outputting K clusters, each cluster corresponding to a drainage pump group, and each drainage pump group including at least one drainage pump; S103, based on each drainage pump group, calculating the state extreme value of each drainage pump in each historical time window to form a state fluctuation sequence of the drainage pump, and generating a state deviation sequence of the drainage pump from each other drainage pump in the drainage pump group; S104, among all state deviation sequences of each drainage pump in the drainage pump group, if there is a state deviation sequence in which the number of historical time windows reaching the deviation threshold is greater than a preset number, the corresponding drainage pump is determined as a suspected abnormal drainage pump; S105, based on each suspected abnormal drainage pump, obtaining its multivariate deviation set and its corresponding abnormal probability value every preset period, and generating an abnormal probability value sequence for each suspected abnormal drainage pump; S106, based on each suspected abnormal drainage pump, generate a hit rate of the abnormal probability value sequence according to a preset abnormal probability threshold and abnormal probability value sequence, and if the hit rate is greater than the hit threshold, determine that the suspected abnormal drainage pump is confirmed to be abnormal.

2. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The vibration characteristic data includes vibration frequency, amplitude, and vibration duration; the working characteristic parameters include flow rate, pressure, rotation speed, and temperature; and the multivariate state characteristic vector of each drainage pump is constructed, including: A1. Calculate the statistics of vibration characteristic data, including the mean, variance, and kurtosis values ​​corresponding to vibration frequency, amplitude, and vibration duration respectively; A2. Calculate the statistics of the working characteristic parameters, including the mean, variance, and kurtosis values ​​corresponding to flow, pressure, speed, and temperature respectively; A3. Combine the statistics of the vibration characteristic data and the statistics of the working characteristic parameters to generate a multivariate state characteristic vector of each drainage pump.

3. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The pre-set clustering algorithm is set to a K-Means clustering algorithm, and the center point of each cluster is determined as the label feature vector of the corresponding drainage pump group.

4. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The state of the drainage pump is Qk], and the state deviation sequence is expressed as [ΔQ1, ΔQ2, ..., ΔQk]; Qk is the state extreme value in the kth historical time window within the preset historical period, and ΔQk is the state deviation value of the drainage pump from other drainage pumps in the drainage pump group.

5. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The calculation method of the state extreme value is: The vibration extreme value and working extreme value corresponding to the vibration characteristic data and working characteristic parameters of each drainage pump in each historical time window are obtained, and the vibration extreme value and the working extreme value are weightedly summed to obtain the state extreme value.

6. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The step of obtaining the multivariate state deviation set and its corresponding abnormal probability value every preset period includes: B1. Based on each suspected abnormal drainage pump, obtain the multivariate state feature vector in the previous cycle and the multivariate state feature vector set of all other drainage pumps in the drainage pump group at every preset cycle; B2. Calculate the Euclidean distance value between the drainage pump and each multivariate state feature vector in the multivariate state feature vector set respectively to form a multivariate state deviation set of the drainage pump. Each element in the set corresponds to a Euclidean distance value. The ratio of the number of elements in the multivariate state deviation set that are greater than a preset distance threshold to the total number of elements is determined as the abnormal probability value.

7. The intelligent control method for building energy-saving water supply and drainage system according to claim 1, characterized in that: The hit rate of the abnormal probability value sequence is set to: the number of abnormal probability values ​​greater than the abnormal probability threshold / the total number of abnormal probability values ​​in the abnormal probability value sequence.

8. The intelligent control method for building energy-saving water supply and drainage system according to claim 7, characterized in that: The S106 includes: when the hit rate of the abnormal probability value sequence is within a preset first interval, the suspected abnormal drainage pump is determined as the target drainage pump, and the suspected abnormal drainage pump at the corresponding point is re-inspected for abnormality using a preset abnormal re-test mechanism, and the final diagnosis result is output; wherein the first interval is set to: [30%, 90%].

9. The intelligent control method for building energy-saving water supply and drainage system according to claim 8, characterized in that: The pre-set abnormal retest mechanism specifically includes: S201, obtaining the Euclidean distance value of the multivariate state feature vector of the target drainage pump and other drainage pumps diagnosed as normal in the drainage pump group in which the target drainage pump is located within a recent preset period; S202, inputting the Euclidean distance value into a pre-trained state abnormality prediction model, and outputting a diagnosis result of the target drainage pump, the diagnosis result including abnormality and normality; Among them, each drainage pump group corresponds to a state abnormality prediction model. The state abnormality prediction model corresponding to the target drainage pump is determined according to the drainage pump group where the target drainage pump is located. The specific process of obtaining the pre-trained state abnormality prediction model is as follows: In a drainage pump group, a large number of multivariate state feature vectors of all drainage pumps in historical periods are obtained, and the drainage pumps with long-term stable and normal operation in the drainage pump group are selected as reference objects. The Euclidean distance values ​​of the multivariate state feature vectors of each drainage pump and the corresponding reference object in all historical periods are obtained as training data sets, and each training data is labeled with a label indicating whether the state of the drainage pump is abnormal or normal; the pre-set neural network structure is trained using the training data set, and the model parameters are continuously optimized to obtain the final state abnormality prediction model.

10. The intelligent control method for building energy-saving water supply and drainage system according to claim 9, characterized in that: Each of the drainage pump groups corresponds to a state abnormality prediction model, each state abnormality prediction model includes a plurality of differentiated prediction models, and S202 further includes: S301, based on the preset first interval, dividing into three sub-intervals [30%, 50%), [50%, 70%), and [70%, 90%), each sub-interval corresponding to a differentiated prediction model; S302, obtaining the hit rate corresponding to the abnormal probability value sequence of the target drainage pump, and locating the corresponding differentiated prediction model according to the sub-interval where the hit rate is located; S303: Input the Euclidean distance value corresponding to the target drainage pump into the corresponding differential prediction model, and output the diagnosis result of the target drainage pump.

Images