Segmentation method and device for secondary water supply flow characteristic curve

Abstract

The invention provides a segmentation method and device for a secondary water supply flow characteristic curve and a water pump unit selection method based on the segmentation method, and the segmentation method comprises the following steps: a data collection step: obtaining water consumption data of a secondary water supply system, and constructing a time sequence flow characteristic curve of the water consumption data; and an optimal segmentation step: dividing the time sequence flow characteristic curve into k characteristic segments by using an optimal segmentation algorithm, and determining an optimal segmentation scheme. The optimal segmentation step further comprises the following steps: a first division step: dividing the time sequence flow characteristic curve by adopting an optimal segmentation algorithm; in the second division step, the maximum fluctuation cumulative deviation value VL-max under the minimum segment length L serves as a measurement standard and is compared with the fluctuation cumulative deviation value vl of each feature segment, if the fluctuation cumulative deviation value vl of a certain feature segment is larger than the maximum fluctuation cumulative deviation value VL-max, the feature segment is independently extracted, and if the fluctuation cumulative deviation value vl of the feature segment is larger than the maximum fluctuation cumulative deviation value VL-max, the maximum fluctuation cumulative deviation value VL-max is not larger than the maximum fluctuation cumulative deviation value VL-max; and further dividing the feature segment by adopting an optimal segmentation algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of flow analysis of secondary water supply systems, and in particular to a method and device for segmenting a secondary water supply flow characteristic curve, and a method for selecting a water pump unit based on the segmentation of the secondary water supply flow characteristic curve. Background Art

[0002] In the secondary water supply system, it is necessary to predict the water supply volume and control the operation of the water supply equipment based on the predicted data.

[0003] The water supply flow of the secondary pump station of the water plant or the secondary water supply flow of the building has obvious periodic changes and has a variety of different flow operating conditions. However, due to the lack of quantitative analysis methods for its segmentation characteristics, many projects directly define a fixed period as the maximum flow section during the day or the period when the low flow condition occurs at night, and only use this as the basis for judging and controlling "day-night segmented water supply", which is no longer suitable for energy-saving requirements. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and device for segmenting a secondary water supply flow characteristic curve, which effectively segments the characteristic curve based on an optimal segmentation algorithm and can be used to improve the design parameters of the secondary water supply system, guide the selection of pump groups, and optimize the control of system operation.

[0005] To achieve the above-mentioned purpose, the present invention provides a method for segmenting a secondary water supply flow characteristic curve, comprising the following steps: a data collection step, obtaining water consumption data of a secondary water supply system, and constructing a time series flow characteristic curve of the water consumption data; an optimal segmentation step, using an optimal segmentation algorithm to divide the time series flow characteristic curve into k characteristic segments, and determining an optimal segmentation scheme, wherein k ≥ 2 and is a positive integer; the optimal segmentation step further comprises: a first segmentation step, using an optimal segmentation algorithm to segment the time series flow characteristic curve; a second segmentation step, using the maximum fluctuation cumulative deviation value V under the minimum segment length L to determine the optimal segmentation scheme; L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l Greater than the maximum fluctuation cumulative deviation value V L-max , then the feature segment is extracted separately, and the feature segment is further divided using the optimal segmentation algorithm.

[0006] The minimum segment length L is determined by the minimum segment duration T required for the operation control of the secondary water supply system and the collection interval duration e of the water use data, as shown in the following formula: Among them, the minimum segment time T is set according to the water supply equipment of the secondary water supply system.

[0007] The water supply equipment is a water pump, and the minimum segment duration T is as follows:

[0008] T=T1+T2+T3

[0009] Among them: T1 is the start-up time of the pump from receiving the pump start signal to the normal operation of the pump; T2 is the time required for the frequency conversion acceleration and deceleration process of the pump; T3 is the stable operation time of the pump.

[0010] The second division step further includes: comparing the duration of the feature segment with 2eL, and if the duration of the feature segment is less than 2eL, the feature segment is no longer further divided.

[0011] The optimal segmentation step also includes: using an improved silhouette coefficient S k To ensure that the minimum segment length is L, when S k = -1, there is a situation where the minimum segment length of the feature segment is less than L, and no iterative calculation is performed at this time; the improved contour coefficient S k The absolute value of the difference between two points is used instead of the average distance to compare the target feature segment with its adjacent feature segments.

[0012] The improved silhouette coefficient S k As follows:

[0013]

[0014] Where: S k represents the average value of the silhouette coefficients of the k feature segments, and -1≤S k ≤1;s i It is expressed as the silhouette coefficient of each water use data, as follows:

[0015]

[0016] Where: i = 1, 2, 3...n, represents each sample point of the water use data; a i represents the average absolute value of the difference between the i-th sample point and other sample points in the feature segment where the sample point is located, b i It represents the minimum value of the average absolute value of the difference between the i-th sample point and all sample points in the adjacent segment.

[0017] The a i The calculation process is as follows:

[0018]

[0019] Wherein, i = 1, 2, 3...n; the water consumption data is a flow set {y1, y2,..., yn},y t Represents the water use data {y1, y2, ..., y n} at time t; the water consumption data {y1,y2,…,y n} is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

[0020] The b i The calculation process is as follows:

[0021]

[0022] Wherein, i=1, 2, 3...n; the water consumption data is a flow set {y1, y2, ..., yn}, and yt represents the water consumption data {y1, y2, ..., y n} at time t; the water consumption data {y1, y2, ..., y n} is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

[0023] The fluctuation cumulative deviation value v l As follows:

[0024] v l =max(v l，t )-min(v l，t )

[0025] Where: v l，t is the accumulated value of the water consumption data fluctuation at time t, where y t represents water consumption data {y1, y2, ..., y n}The water consumption at time t; Indicates the average value of the characteristic segment where the water consumption is located; i l ≤t<i l+1 , l = 1, 2, ..., k-1, and i0 = 1, i k -1=n.

[0026] The maximum fluctuation cumulative deviation value V L-maxAfter determining the minimum segment length L, calculate the first fluctuation cumulative deviation values ​​corresponding to L adjacent water use data at each water use time t, and then compare the maximum values ​​obtained by these first fluctuation cumulative deviation values, where L≤t≤n and t is an integer.

[0027] On the other hand, the present invention also provides a segmentation device for a secondary water supply flow characteristic curve, comprising a processor, the processor being used to execute the following program modules stored in a memory: a data acquisition module, used to obtain water consumption data of a secondary water supply system, and construct a time series flow characteristic curve of the water consumption data; an optimal segmentation module, used to divide the time series flow characteristic curve into k characteristic segments using an optimal segmentation algorithm, and determine an optimal segmentation scheme, wherein k ≥ 2 and is a positive integer; the optimal segmentation module further comprises: a first segmentation unit, used to divide the time series flow characteristic curve using an optimal segmentation algorithm; a second segmentation unit, used to divide the time series flow characteristic curve using a maximum fluctuation cumulative deviation value V under a minimum segment length L L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l Greater than the maximum fluctuation cumulative deviation value V L-max , then the feature segment is extracted separately, and the feature segment is further divided using the optimal segmentation algorithm.

[0028] On the other hand, the present invention also provides a method for selecting a pump unit based on the segmentation of the secondary water supply flow characteristic curve, comprising: using the above-mentioned segmentation method of the secondary water supply flow characteristic curve to determine the optimal segmentation scheme; a pump unit selection step, according to the optimal segmentation scheme, combined with the performance parameters of the pump unit, select a pump unit that matches the optimal segmentation scheme.

[0029] The number of water pumps running simultaneously in the water pump unit shall not exceed 3.

[0030] The water pump unit consists of 4 water pumps, and the water supply flow ratio of the 4 water pumps is 1:1:0.5:0.09.

[0031] In another embodiment, the water pump unit consists of 4 water pumps, including 3 main pumps and 1 small-flow pump, the flow rates of the 3 main pumps are 50%, 50%, and 25% of the design flow rate of the secondary water supply system, respectively, the flow rate of the 1 small-flow pump is 5% of the design flow rate of the secondary water supply system, and the total flow rate of the 4 water pumps is 1.3 times the design flow rate of the secondary water supply system.

[0032] It can be seen from the above scheme that the advantages of the present invention are:

[0033] The method of the present invention effectively divides the secondary water supply flow characteristic curve through the optimal segmentation algorithm, uses the first division to identify the "macro" change trend in the water supply flow, and uses the second division to identify the "micro" change trend in the water supply flow. It can be used to improve the design parameters of the secondary water supply system, guide the selection of pump groups, and optimize the control of system operation.

[0034] The present invention proposes concepts such as minimum segment length and cumulative deviation value of water use fluctuation according to the control requirements of the secondary water supply system operation, so that the segmentation results are more scientific and detailed, and can effectively identify the specific time when the secondary water supply system changes.

[0035] The water supply capacity of the working water pump unit in the water pump unit selection method provided by the present invention not only meets the maximum design flow of the system, but also can simultaneously meet the operating conditions of the small flow section at night, thereby ensuring that the water supply pump unit can operate in the high-efficiency section for a long time. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flow chart of the segmentation method of the secondary water supply flow characteristic curve of the present invention;

[0037] Figure 2 Another flow chart of the segmentation method of the secondary water supply flow characteristic curve of the present invention;

[0038] Figure 3 This is a curve diagram of the monitoring results (L / s) of the water supply flow rate in residential buildings in a specific application scenario of the present invention;

[0039] Figure 4 The present invention adopts Fisher optimal segmentation algorithm to Figure 3 The segmented result curve graph of

[0040] Figure 5 It is a curve diagram of the variation of water supply by a water pump in a specific application scenario of the present invention;

[0041] Figure 6 It is a schematic diagram of the structure of the segmented device of the secondary water supply flow characteristic curve of the present invention;

[0042] Wherein, the reference numerals are:

[0043] 1- Segmentation method of secondary water supply flow characteristic curve;

[0044] 2- Segmentation device of secondary water supply flow characteristic curve;

[0045] 20-processor;

[0046] 21- memory;

[0047] 210-data acquisition module;

[0048] 211-optimal segmentation module;

[0049] 2110-first division unit;

[0050] 2111-second division unit;

[0051] S10, S20, S200, S201 - steps. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments to further understand the purpose, solution and effect of the present invention, but it is not intended to limit the scope of protection of the claims attached to the present invention.

[0053] References to "an embodiment", "another embodiment", "this embodiment", etc. in the specification refer to the embodiment being described which may include specific features, structures or characteristics, but not every embodiment must include these specific features, structures or characteristics. In addition, such expressions do not refer to the same embodiment. Furthermore, when describing specific features, structures or characteristics in conjunction with an embodiment, whether or not there is an explicit description, it is indicated that incorporating such features, structures or characteristics into other embodiments is within the knowledge of those skilled in the art.

[0054] Certain words are used in the specification and subsequent claims to refer to specific components or parts. People of ordinary skill in the art should understand that technical users or manufacturers can refer to the same component or part with different nouns or terms. This specification and claims do not use differences in names as a way to distinguish components or parts, but use differences in the functions of components or parts as the criteria for distinction. "Including" and "including" mentioned throughout the specification and claims are open-ended terms and should be interpreted as "including but not limited to". In addition, the word "connected" here includes any direct and indirect electrical connection means. Indirect electrical connection means include connection through other devices.

[0055] It should be noted that, in the description of the present invention, the terms "lateral", "longitudinal", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and "about", or "approximately", "substantially", "left and right" and the like indicating directions or positional relationships or parameters are all based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description content, and do not indicate or imply that the referred device or element must have a specific direction, specific size, or be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0056] The present invention provides a segmentation method for a secondary water supply flow characteristic curve, which solves the problem in the prior art that there is a lack of a method for quantitatively analyzing segmentation characteristics and the energy-saving requirement is not met.

[0057] In order to solve the above problems, the overall idea of ​​the technical solution adopted by the present invention is as follows: a flow meter is set on the water outlet main of the water supply system, the time series data is read at a frequency of 2 flow data per second, a time series flow characteristic curve is drawn, and the time series flow characteristic curve is divided multiple times using the Fisher optimal segmentation algorithm to determine the optimal segmentation scheme.

[0058] Embodiment 1:

[0059] Figure 1 and Figure 2 A flow chart of a segmentation method for a secondary water supply flow characteristic curve provided in Embodiment 1 of the present invention (hereinafter referred to as Method 1), wherein Method 1 comprises the following steps:

[0060] S10: a data collection step, obtaining water use data of the secondary water supply system and constructing a time series flow characteristic curve of the water use data;

[0061] S20: an optimal segmentation step, using an optimal segmentation algorithm to segment the time series flow characteristic curve into k characteristic segments, and determining an optimal segmentation scheme.

[0062] In this embodiment, the optimal segmentation step S20 further includes:

[0063] S200: a first division step, using an optimal segmentation algorithm to divide the time series flow characteristic curve;

[0064] S201: The second division step is to use the maximum fluctuation cumulative deviation value V under the minimum segment length L L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l >V L-max , then the feature segment is extracted separately, and the optimal segmentation algorithm is used to further divide the feature segment.

[0065] In step S10:

[0066] The flow rate of the secondary water supply system is read at no less than 2 data per second. The time series flow curves of different time intervals such as 1s, 1min, and 5min can be obtained by merging the water use data of the same period. The data is processed as a cycle of 24 hours a day, weekdays, Saturdays and Sundays, and holidays. Through the study of water use events, it is found that the change characteristics of water use data have both obvious discreteness and certain aggregation. Among them, industrial water use is affected by production processes, while the changes in residential water use are the specific reflection of water use events, which are affected by both natural and human factors. For a single water use event, the number of times each water use equipment and sanitary ware is used in a day, the duration of water use, and the time period of occurrence are all random, affected by natural factors such as region and climate and human factors such as user age, gender, and living habits. After investigation and research, it was found that the average water use time for men using washbasins is generally 8.8s, the average water use time for women is about 13.8s, and the shortest water use time is only 2s; the flushing time of the induction urinal is about 3s before use and about 8s after use; the water inflow time of the flush toilet is determined by the water supply pressure and the water tank volume, generally about 10s; the average water use time of residents for showers is about 760s. There are many factors that affect the occurrence of water use events and they are coupled with each other. The probability distribution of water use events in different buildings is completely different. On the time coordinate, the combination of different dates, different time periods, and different water use events is complex and changeable. The building water use data at each moment will undergo huge fluctuations. The smaller the interval for collecting water use data, the more detailed the building water use law it reflects, and the richer the water use event information it contains.

[0067] Therefore, the second flow rate variation characteristics of the urban water supply system and the building water supply system are similar. To scientifically, clearly and reasonably divide the operating conditions of the secondary water supply system, it is necessary to deeply explore the secondary water supply time series characteristic flow curve. The water use data used in this invention is the second flow rate of building water supply.

[0068] In step S20:

[0069] The optimal segmentation algorithm is the Fisher optimal segmentation algorithm. Under the premise of not changing the time arrangement order of water use data, the Fisher optimal segmentation algorithm determines the segmentation scheme of water use data by defining the class diameter and taking the minimum sum of square deviations of each segment after segmentation as the objective function, as shown in formula (1).

[0070]

[0071] Where: D(i l ,i l+1 -1) is the class diameter, i.e. the sum of squared deviations, which indicates the degree of difference between samples of the same type. The larger the sum of squares of sample deviations, the higher the degree of dispersion of the data. n} is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

[0072] Since the Fisher optimal segmentation algorithm takes the minimum sum of squared deviations of each segment as its objective function, it can only identify and extract the "macro" change trend of the building water supply flow rate per second, while the "micro" change trend is ignored.

[0073] Therefore, the optimal segmentation step S20 of the present invention includes the first segmentation step S200 and the second segmentation step S201. Specifically, firstly, the characteristic curve of the building water supply flow rate is "macroscopically" segmented using the Fisher optimal segmentation algorithm, i.e., the first segmentation step S200; then, the maximum fluctuation cumulative deviation value V under the minimum segment length L is used. L-max As a measurement standard, the cumulative deviation value v from the fluctuation of each water use characteristic segment k After comparison, identification and extraction of water usage characteristic segments with "micro" change trends, the dynamic Fisher optimal segmentation algorithm is used to further divide the water usage characteristic segments with "micro" change trends, that is, the second division step S201, and finally the "macro" and "micro" change trends are identified and divided, and the optimal segmentation scheme of the building water supply second flow characteristic curve is obtained.

[0074] In step S201:

[0075] The minimum segment length L of the building water supply flow rate in seconds is determined by the minimum segment length T of the water use characteristic segment (i.e., k characteristic segments) and the collection interval length e of the building water supply flow rate in seconds, as shown in formula (2).

[0076]

[0077] Where: T is the minimum segment duration of the water use characteristic segment, min; e is the duration of the building water supply flow rate collection interval, min.

[0078] The minimum segmentation time T is set according to the water supply equipment (such as a water pump) of the secondary water supply system. In this embodiment, the minimum segmentation time T is mainly related to the operation control level of the water pump, and should not be less than 15 minutes. It is not economical to operate less than 15 minutes. The maximum operation time is determined according to the water use characteristics. In the process of achieving a two-way balance between water use and water supply in the building water supply system, the water pump unit needs to complete the start-up instruction to start working and output flow, and it needs to complete the pressure sensor after receiving the signal to make an adjustment instruction to adjust the water flow rate of the water pump. Therefore, within the minimum segmentation time T, the secondary water supply system needs to complete the start-up and adjustment process of the water pump and be able to maintain stable operation for a long period of time, as shown in formula (3).

[0079] T=T1+T2+T3(3)

[0080] Where: T1 is the pump start time from receiving the pump start signal to the normal operation of the pump, min; T2 is the time required for the pump frequency conversion acceleration and deceleration process, min; T3 is the stable operation time of the pump, min. The minimum segment time T of the building water supply flow rate in seconds ensures that the pump can complete the operation control instructions within the specified time and ensure that the pump can operate smoothly for a long period of time.

[0081] Therefore, the minimum segment length L not only meets the requirements of stable operation of the secondary water supply system, avoids the impact on the system when the water pump is started at power frequency, and prolongs the service life of the water pump. Moreover, this value can ensure the stable operation of the water pump for a long period of time, providing an operation control basis for achieving stable operation of the water pump in the high-efficiency section.

[0082] In order to ensure that the optimal segmentation scheme meets the setting requirements of the minimum segmentation length L, the present invention improves the contour coefficient method so that the final segmentation scheme can meet the requirements of water use characteristic segmentation. The contour coefficient method combines cohesion and separation as a measure of clustering effect. In ordered clustering, there is a fixed order between the divided water use characteristic segments, and there is no need to compare the target characteristic segment with all other characteristic segments when calculating the difference between classes. The improved contour coefficient method uses the absolute value of the difference between two points instead of the average distance, and compares the target characteristic stage with its adjacent characteristic stage, so as to determine the optimal segmentation scheme for the building water supply second flow rate. The calculation process is shown in formula (4).

[0083]

[0084] Where: S k It represents the average value of the improved silhouette coefficient when the number of segmentation segments is k, and the improved silhouette coefficient S k The value varies between -1 and 1, that is, -1≤S k ≤1, S kThe larger the value, the better the solution with k segments. i It is expressed as the improved silhouette coefficient of each water use data, and its calculation process is shown in formula (5).

[0085]

[0086] Where: a i represents the average absolute value of the difference between the i-th object and all other objects in the segment. Its calculation process is shown in formula (6); b i It represents the minimum value of the average absolute value of the difference between the i-th object and all objects in the adjacent segment. Its calculation process is shown in formula (7).

[0087]

[0088] Wherein, i = 1, 2, 3...n; the water consumption data is a flow set {y1, y2,..., y n},y t Represents the water use data {y1, y2, ..., y n} at time t; the water consumption data {y1,y2,…,y n} is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

[0089] When the number of data points in a certain feature segment is less than L, the improved silhouette coefficient S of the segmentation scheme is defined as k = -1, as the penalty coefficient. k = -1, the minimum segment length of the feature segment is less than L, and no iterative calculation is performed at this time.

[0090] Since the Fisher optimal segmentation algorithm starts from the sum of squared deviations of the overall data, it is difficult to effectively identify the "micro" change trend that exists locally in the second flow of building water supply. In order to effectively identify whether there is a micro change trend in each water use characteristic segment, the fluctuation cumulative deviation value is proposed for judgment. First, the water use fluctuation of each characteristic segment is extracted, as shown in formula (8).

[0091]

[0092] In the formula, y t represents the building water supply flow rate in seconds {y1,y2,…,y n}The water consumption at time t; It represents the average value of all water consumption in the water consumption characteristic segment where the water consumption is located. There are large differences in water consumption fluctuations in different water consumption characteristic stages. The water consumption fluctuations in water consumption characteristic segments with "micro" change trends cannot be evenly distributed around 0. In order to measure the unevenness of fluctuations in different water consumption characteristic stages, the segments with micro-change trends are effectively identified, and the fluctuation cumulative deviation value v of the characteristic segment is defined l , as shown in formula (9).

[0093] v l =max(v l,t )-min(v l,t ) (9)

[0094] Where: v l,t is the accumulated value of water consumption fluctuation at time t, where i l ≤t l+1 ,l=1,2……,k-1, and i0=1,i k -1=n. Volatility cumulative deviation value v l The larger it is, the more uneven the water use fluctuations within the water use characteristic section are.

[0095] At the same time, the maximum fluctuation cumulative deviation value V L-max There is a relationship between the minimum segment length L and the minimum segment length L. If the minimum segment length of the water-use characteristic segment is L, that is, the water supply flow rate of L adjacent buildings {y t-L ,y t-L+1 ,…,y t} (where L≤t≤n and t is an integer) is divided into a water use characteristic segment. After determining the minimum segment length, without segmenting the water use data, according to formula (8) and formula (9), the first fluctuation cumulative deviation value v corresponding to the L adjacent building water supply flow rates at each water use time is divided into a water use characteristic segment. t , the first fluctuation cumulative deviation value v of a total of L water use data t , and then compare these first fluctuation cumulative deviation values ​​v t Get the maximum fluctuation cumulative deviation value V L-max , the maximum fluctuation cumulative deviation value V L-max The cumulative deviation value v of the fluctuation of each characteristic segment after segmentation l Specifically, when the minimum segment length is set to L, if the cumulative deviation v of a certain water use characteristic segment is l >V L-max ​, indicating that the feature segment has a microscopic change trend, so the water use feature segment is extracted separately. First, the duration of the water use feature segment is compared with 2eL. If the duration of the water use feature segment is less than 2eL, then according to the minimum segmentation requirement, the water use feature segment cannot be further segmented.

[0096] Only when this water consumption characteristic segment satisfies the fluctuation cumulative deviation v l >V L-max When the duration is greater than 2eL, the water feature segment is extracted and further segmented until S k = -1, it indicates that at this time, any segmentation scheme of the water feature segment must have a minimum segment length less than L, and then no iterative calculation will be performed.

[0097] To facilitate understanding of the above embodiment, a specific application scenario of the above embodiment will be used as an example for description below:

[0098] Embodiment 2:

[0099] The water consumption data verified in this application scenario is the measured building water supply flow rate in seconds in a residential community. There are 3 multi-storey residential buildings in the community, with the highest building having 7 floors. The community has a high occupancy rate, with a total of 70 households and a total of about 245 residents. During the monitoring period, the highest daily domestic water consumption in the community occurred on January 21, 2023 (the 30th day of the twelfth lunar month, New Year's Eve), with a daily domestic water consumption of 32.28m 3 The collection interval of water use data is only 1 minute, and the number of data collected in one day is 1440. With the shortening of the collection interval, the amount of water use data collected has doubled, and the fluctuation characteristics of water use data have become more obvious, which has increased the difficulty of analyzing the trend of water use changes. Figure 3 shown.

[0100] Before formally performing the segmentation calculation, it is also necessary to modify the two iterative judgment parameters of the dynamic Fisher optimal segmentation algorithm, the cyclic judgment criterion and the improved contour coefficient, according to the data characteristics and the minimum segmentation requirements, to ensure that the segmentation results meet the requirements of stable operation of the secondary water supply equipment.

[0101] The investigation and study of the operating parameters of the water pump in the community found that during the water pump adjustment process, the water pump has just started, the system parameters change greatly, and the automatic start time from receiving the pump start signal to the normal operation of the water pump is generally 2 minutes; during the inertial convergence and stabilization process, the acceleration and deceleration time of the water pump frequency conversion is generally set to 50Hz / 5~10s. In order to ensure the stability of the secondary water supply system, avoid the impact on the system when the water pump is started at the power frequency, and extend the service life of the water pump. Combined with the operation control characteristics of the water pump and the demand for stable water supply of the secondary water supply system, the minimum segmentation time of the water use characteristic segment T=15min is more reasonable.

[0102] The minimum duration of the water use characteristic segment is T = 15 min, and the collection interval of the building water supply flow rate in seconds is e = 1 min. According to formula (2), the minimum segment length of the building water supply flow rate in seconds is L = 15, which means that at least 15 adjacent building water supply flow rates are allowed to be divided into one segment.

[0103] According to equations (8) and (9), when the minimum segment length of the building water supply flow rate is L = 15, the fluctuation cumulative deviation value v corresponding to each water use moment is t The maximum value in is 127.33L, so the cycle determination basis of the dynamic Fisher optimal segmentation algorithm is V 15-max =127.33L.

[0104] The segmentation method 1 of the secondary water supply flow characteristic curve provided by the present invention is used to calculate the segmentation result of the building water supply flow rate in seconds, such as Figure 4 shown.

[0105] Specifically, the start and end time of each water use characteristic segment, the average water consumption of each stage, and the corresponding cumulative deviation value of each stage and other related parameters are shown in Table 1.

[0106] Table 1 Duration and cumulative deviation of each water use characteristic segment based on Fisher optimal segmentation algorithm

[0107]

[0108]

[0109] like Figure 4, the segmentation results of the building water supply second flow rate by Fisher optimal segmentation algorithm show that: the longest water use characteristic segment occurs from 1:00 to 4:28, lasting for 208 minutes. During this water use characteristic stage, the residents of the community are in a state of rest, and water-using appliances and equipment are basically rarely used, and the average hourly water consumption of this characteristic segment is only 110.77L / h at the minimum; the shortest water use characteristic segment occurs from 10:43 to 10:58, lasting for only 15 minutes. The average hourly water consumption of this water use characteristic segment is significantly higher than that of the two adjacent water use characteristic segments, and the average hourly water consumption of the two adjacent water use characteristic segments is not much different. This phenomenon shows that the present invention can effectively identify the "micro" change trend in water use data; the water use characteristic segment with the largest average hourly water consumption occurs from 19:43 to 21:59, lasting for 136 minutes. At the same time, due to the high degree of discreteness of the water use data of this characteristic segment, the segment cannot be further segmented, and the fluctuation cumulative deviation value of this characteristic segment reaches 365.38L.

[0110] The number of segments into which water use data is divided is related to the changing characteristics of the building water supply second flow rate itself, and there is no strict regulation. Therefore, the present invention proposes concepts such as the minimum segment length and the cumulative deviation value of water use fluctuations based on the control requirements of the operation of the secondary water supply equipment, so that the segmentation results are more scientific and detailed, and can effectively identify the specific moment when the building water supply second flow rate changes, and reflect the rising, peak and falling trend characteristics of the small flow segment at night and the various water use peak periods during the day. The building water supply second flow rate segmentation method proposed in the present invention provides a scientific and reasonable data mining tool for in-depth research and analysis of the changing characteristics of the building water supply second flow rate and the optimized operation control of the secondary water supply system.

[0111] Embodiment three:

[0112] When selecting water pumps for water supply system design, the prior art often combines water pumps based on the maximum design flow of the water supply system. In actual projects, there are multiple water pumps running in parallel, such as one for use and one for backup, two for use and one for backup, three for use and one for backup, and four for use and one for backup. The flow matching includes 50%+50%, 40%+40%+20%, 10%+30%+30%+30%, 25%+25%+25%+25%, etc. This method often cannot effectively take into account the low-flow operation conditions at night, causing the water supply system to operate in a non-efficient section for a long time.

[0113] In the actual application scenario of the second embodiment, the present invention provides a method for selecting a pump unit based on the segmentation method 1 of the secondary water supply flow characteristic curve. As shown in Table 1, the pump unit is selected according to the segmentation result (i.e., the optimal segmentation scheme) and the performance parameters of the pump, so that the pump operates in the high-efficiency section. In this embodiment, the pump unit is 4 pumps, namely A+2B+C, wherein: A is a small flow pump with a water supply flow of 110L / h and 1 unit; B is a large flow pump with a water supply flow of 1250L / h and 2 units; C is a medium flow pump with a water supply flow of 650L / h and 1 unit. For example: when the water consumption is relatively small, only the A small flow pump can be turned on; when the water consumption increases, the C or B flow pump can be selected, or any combination of the three pumps, but the number of pumps running at the same time does not exceed 3. In this way, most of the secondary water supply system operates in the high-efficiency section, and the secondary water supply system is economical and reasonable. The amplitude of change in water supply of different pumps is also different, such as Figure 5 shown.

[0114] Specifically: Combine the pump operation characteristic curve QH to determine the high-efficiency section Q1~Q3 of the pump operation, among which the pump operation efficiency η2 at the operating point Q2 is the highest. Select the pump whose design operating point Q2 is consistent with the average water consumption of the water use characteristic segment, so as to ensure that the pump can operate in the high-efficiency section Q1~Q3.

[0115] This embodiment uses the segmentation method 1 of the secondary water supply flow characteristic curve to perform segmented analysis on the water supply characteristic curve, and provides a decision-making basis for the selection of the water pump in the small flow section at night. Specifically, the water supply flow ratio of the four water pumps of the water pump unit can be selected according to the proportional relationship criterion of 1:1:0.5:0.09. Exemplarily, the water pump unit is composed of three main pumps of 50%, 50%, and 25% of the maximum design flow of the secondary water supply system, and one small flow pump of 5% of the maximum design flow, and the number of water pumps put into operation throughout the day does not exceed 3. At this time, the total flow of the water pump unit is 1.3 times the design flow of the secondary water supply system.

[0116] The advantage of this embodiment is that the water supply capacity of the working water pump unit not only meets the maximum design flow of the secondary water supply system, but also can simultaneously meet the operating conditions of the small flow section at night, thereby ensuring that the water supply pump group can operate in the high-efficiency section for a long time.

[0117] The following is a device embodiment corresponding to the above method embodiment. This embodiment can be implemented in conjunction with the above embodiment. The relevant technical details mentioned in the above embodiment are still valid in this embodiment. In order to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied in the above embodiment.

[0118] Embodiment 4:

[0119] Embodiment 4 of the present invention further provides a segmentation device 2 for a secondary water supply flow characteristic curve, comprising a processor 20, wherein the processor 20 is used to execute the following program modules stored in a memory 21:

[0120] The data acquisition module 210 is used to obtain the water consumption data of the secondary water supply system and construct a time series flow characteristic curve of the water consumption data;

[0121] The optimal segmentation module 211 is used to divide the time series flow characteristic curve into k characteristic segments using an optimal segmentation algorithm and determine an optimal segmentation scheme, wherein k≥2 and is a positive integer.

[0122] In this embodiment, the optimal segmentation module 211 further includes: a first segmentation unit 2110, which is used to segment the time series flow characteristic curve using the optimal segmentation algorithm; a second segmentation unit 2111, which is used to divide the maximum fluctuation cumulative deviation value V under the minimum segment length L L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l Greater than the maximum fluctuation cumulative deviation value V L-max , then the feature segment is extracted separately, and the feature segment is further divided using the optimal segmentation algorithm.

[0123] The data acquisition module 210 is used to read the flow of the secondary water supply system at no less than 2 data per second, and form a time series flow curve through time series flow processing, and process the data according to 24 hours a day, weekdays, weekends and holidays as a cycle. The water use data used in this embodiment is the building water supply second flow.

[0124] The optimal segmentation algorithm in the optimal segmentation module 211 is the Fisher optimal segmentation algorithm. Under the premise of not changing the time arrangement order of the water use data, the Fisher optimal segmentation algorithm determines the segmentation scheme of the water use data by defining the class diameter and taking the minimum sum of square deviations of each segment after segmentation as the objective function, as shown in formula (1).

[0125] Since the Fisher optimal segmentation algorithm takes the minimum sum of squared deviations of each segment as its objective function, it can only identify and extract the "macro" change trend of the building water supply flow rate per second, while the "micro" change trend is ignored.

[0126] Therefore, the optimal segmentation module 211 of this embodiment includes a first segmentation unit 2110 and a second segmentation unit 2111. Specifically, the first segmentation unit 2110 is used to perform a "macro" segmentation on the characteristic curve of the building water supply flow rate per second using the Fisher optimal segmentation algorithm; the second segmentation unit 2111 is used to perform a "macro" segmentation on the characteristic curve of the building water supply flow rate per second using the maximum fluctuation cumulative deviation value V under the minimum segment length L. L-max As a measurement standard, the cumulative deviation value v from the fluctuation of each water use characteristic segment l After comparing and identifying the water use characteristic segments with "micro" change trends and extracting them, the dynamic Fisher optimal segmentation algorithm is used to further divide the water use characteristic segments with "micro" change trends. Finally, the "macro" and "micro" change trends are identified and divided, and the optimal segmentation scheme of the building water supply second flow characteristic curve is obtained.

[0127] In the second dividing unit 2111:

[0128] The minimum segment length L of the building water supply flow rate in seconds is determined by the minimum segment length T of the water use characteristic segment (i.e., k characteristic segments) and the collection interval length e of the building water supply flow rate in seconds, as shown in formula (2).

[0129] The minimum segmentation time T is set according to the water supply equipment (such as a water pump) of the secondary water supply system. In this embodiment, the minimum segmentation time T is mainly related to the operation control level of the water pump, and should not be less than 15 minutes. It is not economical to operate less than 15 minutes. The maximum operation time is determined according to the water use characteristics. In the process of achieving a two-way balance between water use and water supply in the building water supply system, the water pump unit needs to complete the start-up instruction to start working and output flow, and it is necessary to complete the pressure sensor to make an adjustment instruction to adjust the water flow after receiving the signal. Therefore, within the minimum segmentation time T, the secondary water supply system needs to complete the start-up and adjustment process of the water pump and can maintain stable operation for a long period of time, as shown in formula (3). The minimum segmentation time T of the building water supply second flow rate not only ensures that the water pump can complete the operation control instruction within the specified time, but also ensures that the water pump can operate smoothly for a long period of time.

[0130] Therefore, the minimum segment length L not only meets the requirements of stable operation of the secondary water supply system, avoids the impact on the system when the water pump is started at power frequency, and prolongs the service life of the water pump. Moreover, this value can ensure the stable operation of the water pump for a long period of time, providing an operation control basis for achieving stable operation of the water pump in the high-efficiency section.

[0131] In order to ensure that the optimal segmentation scheme meets the setting requirements of the minimum segmentation length L, the present invention improves the contour coefficient method so that the final segmentation scheme can meet the requirements of water use characteristic segmentation. The contour coefficient method combines cohesion and separation as a measure of clustering effect. In ordered clustering, there is a fixed order between the divided water use characteristic segments, and there is no need to compare the target characteristic segment with all other characteristic segments when calculating the difference between classes. The improved contour coefficient method uses the absolute value of the difference between two points instead of the average distance, and compares the target characteristic stage with its adjacent characteristic stage, so as to determine the optimal segmentation scheme for the building water supply second flow rate. The calculation process is shown in equations (4) to (7).

[0132] When the number of data points in a certain feature segment is less than L, the improved silhouette coefficient S of the segmentation scheme is defined as k = -1, as the penalty coefficient. k = -1, the minimum segment length of the feature segment is less than L, and no iterative calculation is performed at this time.

[0133] Since the Fisher optimal segmentation algorithm starts from the overall data, it is difficult to effectively identify the "micro" change trend that exists locally in the second flow rate of building water supply. In order to effectively identify whether there is a micro change trend in each water use characteristic segment, a fluctuation cumulative deviation value is proposed for judgment. First, the water use fluctuation of each characteristic segment is extracted, as shown in formula (8). There are large differences in water use fluctuations in different water use characteristic stages. The water use fluctuations in water use characteristic segments with "micro" change trends cannot be evenly distributed around 0. In order to measure the unevenness of fluctuations in different water use characteristic stages and effectively identify segments with micro change trends, the fluctuation cumulative deviation value v of the characteristic segment is defined. l , as shown in formula (9). The larger the fluctuation cumulative deviation value, the more uneven the water use fluctuation change in the water use characteristic section.

[0134] At the same time, the maximum fluctuation cumulative deviation value V L-max There is a relationship between the minimum segment length L and the minimum segment length L. If the minimum segment length of the water-use characteristic segment is L, that is, the water supply flow rate of L adjacent buildings {y t-L ,y t-L+1 ,…,y t} (where L≤t≤n and t is an integer) is divided into a water use characteristic segment. After determining the minimum segment length, without segmenting the water use data, according to formula (8) and formula (9), the first fluctuation cumulative deviation value v corresponding to the L adjacent building water supply flow rates at each water use time is divided into a water use characteristic segment. t , the first fluctuation cumulative deviation value v of a total of L water use data t , and then compare these first fluctuation cumulative deviation values ​​vt Get the maximum fluctuation cumulative deviation value V L-max , the maximum fluctuation cumulative deviation value V L-max The cumulative deviation value v of the fluctuation of each characteristic segment after segmentation l A comparison is made to draw conclusions on whether there are micro-variation trends in the segmentation scheme.

[0135] Specifically, when the minimum segment length is set to L, if the cumulative deviation v of a certain water use characteristic segment is l >V L-max , indicating that the feature segment has a microscopic change trend, so the water use feature segment is extracted separately. First, the duration of the water use feature segment is compared with 2eL. If the duration of the water use feature segment is less than 2eL, then according to the minimum segmentation requirement, the water use feature segment cannot be further segmented.

[0136] Only when this water consumption characteristic segment satisfies the fluctuation cumulative deviation v l >V L-max When the duration is greater than 2eL, the water feature segment is extracted and further segmented until S k = -1, it indicates that at this time, any segmentation scheme of the water feature segment must have a minimum segment length less than L, and then no iterative calculation will be performed.

[0137] In summary, the present invention stipulates the length of the minimum segment according to the control requirements of the operation of the secondary water supply equipment. The segmentation results are more scientific and detailed, and can effectively identify the specific moment when the building water supply flow rate changes, and reflect the rising, peak and falling trend characteristics of the nighttime small flow segment and the various water consumption peak periods during the day. The present invention proposes a segmentation method, which provides a scientific and reasonable data mining tool for in-depth research and analysis of the changing characteristics of the building water supply flow rate and the optimized operation control of the secondary water supply system.

[0138] In addition, the water pump unit selection method based on the segmentation method 1 of the secondary water supply flow characteristic curve provided by the present invention can make the water supply capacity of the working water pump unit not only meet the maximum design flow of the system, but also meet the operating conditions of the small flow section at night, thereby ensuring that the water supply pump group can operate in the high-efficiency section for a long time.

[0139] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which belong to the protection scope of the present invention.

Claims

1. A segmentation method for a secondary water supply flow characteristic curve, characterized in that: The following steps are involved: A data collection step, obtaining water use data of the secondary water supply system, and constructing a time series flow characteristic curve of the water use data; An optimal segmentation step, using an optimal segmentation algorithm to divide the time series flow characteristic curve into k characteristic segments, and determining an optimal segmentation scheme, wherein k ≥ 2 and is a positive integer; The optimal segmentation step further comprises: The first division step is to divide the time series flow characteristic curve using an optimal segmentation algorithm; The second division step is to use the maximum fluctuation cumulative deviation value V under the minimum segment length L L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l Greater than the maximum fluctuation cumulative deviation value V L-max , then the feature segment is extracted separately, and the feature segment is further divided using the optimal segmentation algorithm.

2. The method according to claim 1, characterized in that The minimum segment length L is determined by the minimum segment duration T and the collection interval duration e of the water use data, as shown in the following formula: Among them, the minimum segment time T is set according to the water supply equipment of the secondary water supply system.

3. The method according to claim 2, characterized in that The water supply equipment is a water pump, and the minimum segment duration T is as follows: T=T1+T2+T3 Where: T1 is the pump start time from receiving the pump start signal to the normal operation of the pump; T2 is the time required for the water pump's frequency conversion acceleration and deceleration process; T3 is the time it takes for the water pump to operate stably.

4. The method according to claim 2, characterized in that: The second dividing step further comprises: The duration of the feature segment is compared with 2eL. If the duration of the feature segment is less than 2eL, the feature segment is not further divided.

5. The method according to claim 2, characterized in that: The optimal segmentation step further comprises: Using the improved silhouette coefficient S k To ensure that the minimum segment length is L, when S k = -1, the minimum segment length of the feature segment is less than L, and no iterative calculation is performed at this time; The improved silhouette coefficient S k The absolute value of the difference between two points is used instead of the average distance to compare the target feature segment with its adjacent feature segments.

6. The method according to claim 5, characterized in that The improved silhouette coefficient S k As follows: Where: S k represents the average value of the silhouette coefficients of the k feature segments, and -1≤S k ≤1; s i It is expressed as the silhouette coefficient of each water use data, as follows: Where: i = 1, 2, 3...n, represents each sample point of the water use data; a i represents the average absolute value of the difference between the i-th sample point and other sample points in the feature segment where the sample point is located, b i It represents the minimum value of the average absolute value of the difference between the i-th sample point and all sample points in the adjacent segment.

7. The method according to claim 6, characterized in that The a i The calculation process is as follows: Wherein, i = 1, 2, 3...n; the water consumption data is a flow set {y1, y2,..., y n },y t Represents the water use data {y1, y2, ..., y n } at time t; the water consumption data {y1,y2,…,y n } is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

8. The method according to claim 6, characterized in that The b i The calculation process is as follows: Wherein, i = 1, 2, 3...n; the water consumption data is a flow set {y1, y2,..., y n },y t Represents the water use data {y1, y2, ..., y n } at time t; the water consumption data {y1,y2,…,y n } is divided into k segments (k ≥ 2), which contains k-1 water use data segmentation points; i l Indicates the location of the water use data segmentation point, l = 1, 2..., k-1, and i0 = 1, i k -1=n.

9. The method according to claim 8, characterized in that The fluctuation cumulative deviation value v l As follows: v l =max(v l,t )-min(v l,t ) Where: v l,t is the accumulated value of the water consumption data fluctuation at time t, where y t Represents water consumption data {y1,y2,…,y n }The water consumption at time t; Indicates the average value of the characteristic segment where the water consumption is located; i l ≤t≤i l+1 -1, l=1,2……,k-1, and i0=1, i k -1=n.

10. The method according to claim 9, characterized in that The maximum fluctuation cumulative deviation value V L-max After determining the minimum segment length L, calculate the first fluctuation cumulative deviation values ​​corresponding to L adjacent water use data at each water use moment, and compare the maximum values ​​obtained by these first fluctuation cumulative deviation values.

11. A segmentation device for a secondary water supply flow characteristic curve, characterized in that: The system comprises a processor configured to execute the following program modules stored in a memory: A data acquisition module, used to obtain water consumption data of the secondary water supply system and construct a time series flow characteristic curve of the water consumption data; An optimal segmentation module, used to divide the time series flow characteristic curve into k characteristic segments using an optimal segmentation algorithm, and determine an optimal segmentation scheme, wherein k ≥ 2 and is a positive integer; The optimal segmentation module further comprises: A first segmentation unit, configured to segment the time series flow characteristic curve using an optimal segmentation algorithm; The second division unit is used to accumulate the deviation value V with the maximum fluctuation under the minimum segment length L. L-max As a measure, the cumulative deviation value v from the fluctuation of each characteristic segment is l For comparison, if the cumulative deviation value v of a certain characteristic segment l Greater than the maximum fluctuation cumulative deviation value V L-max , then the feature segment is extracted separately, and the feature segment is further divided using the optimal segmentation algorithm.

12. A method for selecting a pump unit based on segmented secondary water supply flow characteristic curve, characterized in that: include: Determine the optimal segmentation scheme by using the segmentation method of the secondary water supply flow characteristic curve according to any one of claims 1 to 10; The water pump unit selection step is to select a water pump unit that matches the optimal segmentation scheme according to the optimal segmentation scheme and in combination with the performance parameters of the water pump unit.

13. The method according to claim 12, characterized in that The number of water pumps running simultaneously in the water pump unit shall not exceed 3.

14. The method according to claim 12, characterized in that The water pump unit consists of 4 water pumps, and the water supply flow ratio of the 4 water pumps is 1:1:0.5:0.

09.

15. The method according to claim 12, characterized in that The water pump unit consists of 4 water pumps, including 3 main pumps and 1 small-flow pump. The flow rates of the 3 main pumps are 50%, 50% and 25% of the design flow rate of the secondary water supply system respectively, and the flow rate of the 1 small-flow pump is 5% of the design flow rate of the secondary water supply system. The total flow rate of the water pump unit is 1.3 times the design flow rate of the secondary water supply system.

Images