Water supply network water hammer early warning method and system

Abstract

The invention discloses a water supply pipe network water hammer early warning method and system, relates to the technical field of pipe network safety monitoring, and solves the technical problem that a water hammer phenomenon judgment result is inaccurate due to the fact that multiple analysis methods for adapting to specific application scenes do not exist in the prior art. A water hammer triggering mechanism is determined through a steady-state pressure multiple method, and whether a first water hammer state occurs or not is judged based on the water hammer triggering mechanism; if yes, judging whether a second water hammer state occurs or not through a pressure slope method; performing water hammer phenomenon identification through a long and short time window method; judging whether a pre-water hammer phenomenon occurs or not; obtaining a shortest path through a dynamic time warping algorithm, obtaining warping pressure data through the shortest path, and judging whether a pre-water hammer phenomenon occurs in combination with a square correlation coefficient; judging whether a water hammer event occurs or not according to the early warning signal; according to the method, mutual verification is carried out in the water hammer event judgment through multiple judgment mechanisms, and the judgment accuracy and reliability are improved.

Description

Technical Field

[0001] The invention belongs to the field of pipe network safety monitoring, relates to a water hammer early warning technology, and specifically provides a water hammer early warning method and system for a water supply pipe network. Background Art

[0002] The water hammer monitoring system for water supply pipes is a device that can monitor and record pressure changes in pipes in real time. Through sensors installed on pipes, the system can monitor water flow velocity and pressure changes in pipes in real time. Once an abnormal water hammer signal is detected, the system will immediately issue an alarm and notify relevant personnel to handle it. At the same time, the system will also record the time, location and cause of these abnormal signals to provide a basis for subsequent processing and analysis. It is suitable for real-time monitoring and identification of water hammer in pressurized water delivery projects such as urban water supply.

[0003] The prior art (the invention patent application with the publication number CN115739864A) discloses a water hammer analysis method for CIP cleaning equipment and pipelines, comprising the following steps: obtaining valve information related to the cleaning step, establishing an association model between the cleaning step and the related valve; obtaining a working state feature group of the current cleaning step, wherein the working state feature group of the current cleaning step includes two or more working state features; comparing the working state feature group of the current cleaning step with a corresponding feature threshold to determine whether the water hammer occurrence condition is met; when it is determined that the water hammer occurrence condition is met, determining the suspicious valve causing the water hammer based on the cleaning step causing the water hammer and the association model. By obtaining the working state feature group of the current cleaning step in real time and comparing it with the corresponding feature threshold, it is accurately determined whether the water hammer actually occurs, and then the suspicious valve causing the water hammer is secondary determined; the prior art is applicable to CIP cleaning equipment with a clear interaction between the cleaning step and the valve, but in some complex systems, it is difficult to accurately establish an association model; at the same time, for pipelines of different types and specifications, the prior art does not have a variety of analysis methods to adapt to their specific application scenarios, resulting in the technical problem of inaccurate water hammer phenomenon judgment results.

[0004] The present invention provides a water hammer early warning method and system for a water supply network to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a water hammer early warning method and system for a water supply pipe network, which is used to solve the technical problem that the prior art is difficult to accurately establish an association model in a complex system and does not have a variety of analysis methods to adapt to its specific application scenarios, resulting in inaccurate water hammer phenomenon judgment results.

[0006] To achieve the above objectives, a first aspect of the present invention provides a water hammer early warning method for a water supply network, comprising:

[0007] The instantaneous pressure data collected by the sensor in real time is used to determine the water hammer triggering mechanism through the steady-state pressure multiple method;

[0008] Determine whether the first water hammer state occurs based on the water hammer trigger mechanism;

[0009] If yes, the pressure slope method is used to determine whether the second water hammer state has occurred; if yes, the pre-water hammer phenomenon is marked as occurring and a warning signal 1 is issued; if no, the pre-water hammer phenomenon is marked as not occurring;

[0010] If not, continue to judge;

[0011] Retrieve the water hammer signal that triggers the first water hammer state, and identify the water hammer phenomenon by using the long and short time window method; determine whether the pre-water hammer phenomenon occurs; if yes, issue the second warning signal; if no, mark the pre-water hammer phenomenon as not occurring;

[0012] The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues;

[0013] Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

[0014] Preferably, the determining whether the first water hammer state occurs based on the water hammer trigger mechanism includes:

[0015] Store a number of data according to a set period to construct a pressure series, obtain the median of the pressure data in the corresponding pressure series, and store it in the pressure median series; obtain the mean of the median data in the pressure median series, and use the mean as the steady-state pressure value;

[0016] The trigger judgment mechanism is set as follows: the high pressure threshold is set to 1.06 times the steady-state pressure value, and the low pressure threshold is set to 0.94 times the steady-state pressure value; the pressure data in the pressure series is greater than or equal to the high pressure threshold or the pressure data is lower than the low pressure threshold;

[0017] Determine whether the median sequence is stable; if so, determine in turn whether the pressure data in the pressure sequence reaches the trigger judgment mechanism; if so, mark the first water hammer state as triggered; if not, continue to judge;

[0018] If not, the first water hammer state is marked as not triggered, the median sequence is updated and the judgment is made again.

[0019] The present invention uses the mean of the median data in the median series as the steady-state pressure value, thereby reducing the impact of individual abnormal data on the calculation of the steady-state pressure value and improving the accuracy of monitoring; by judging the stability of the median series, it can ensure that the trigger judgment is performed in a stable pressure environment and avoid false triggering due to pressure fluctuations.

[0020] Preferably, the determining whether the median sequence is stable includes:

[0021] Retrieve the data from the median series and label it as Z k , mark the minimum value in the data of the median sequence as Z min , mark the maximum value in the data of the median sequence as Z max ; Wherein, k is a positive integer;

[0022] By formula Y k =(Z k -Z min ) / (Z max -Z min ) Calculate the normalized standard deviation;

[0023] Determine whether the standard deviations are all less than 0.1; if yes, the median sequence is stable; if no, the median sequence is unstable.

[0024] Preferably, the determining whether the second water hammer state occurs by the pressure slope method includes:

[0025] Retrieve the pressure data corresponding to the first water hammer state trigger, and store the corresponding pressure data in the water hammer sequence until the pressure data in the water hammer sequence reaches the set length; set the set length as the water hammer detection window time length and mark it as hammer width ; and mark the pressure slope threshold as slope, obtain all pressure data within the set time after the trigger, mark the set time as t, and mark the maximum pressure value in all data as P max , the minimum pressure value is marked as P min ; Among them, hammer width >t>0;

[0026] By formula jump value =|P max -P min |Calculate the pressure difference;

[0027] By formula slope press =jump value / hammer width The real-time pressure slope is calculated;

[0028] The time corresponding to the maximum pressure value is marked as t1, and the time corresponding to the minimum pressure value is marked as t2;

[0029] Set the time difference threshold and calculate the time difference using the formula Δt=|t1-t2|;

[0030] Determine the slope press Is it greater than the slope?

[0031] If yes, it is determined whether Δt is greater than the time difference threshold; if yes, the second water hammer state is marked as occurring; if no, the second water hammer state is marked as not occurring;

[0032] If not, the second water hammer state is marked as not occurring.

[0033] It should be noted that, regardless of whether the second water hammer state occurs, it is necessary to reset the first and second water hammer states, reset the median series and the water hammer series, and re-judge.

[0034] By setting the water hammer detection window time length and the pressure slope threshold, the present invention can more accurately capture the characteristics of the water hammer phenomenon, thereby reducing the possibility of false alarms and missed alarms; by calculating the pressure difference and the real-time pressure slope, and combining the time difference corresponding to the maximum pressure value and the minimum pressure value, the authenticity of the water hammer state can be further verified, thereby improving the accuracy of detection.

[0035] Preferably, the water hammer phenomenon identification by long and short time window method includes:

[0036] After the trigger mechanism is implemented, the characteristic function value of the water hammer signal at time m is obtained and marked as x(m); the length of the short time window is marked as ns, and the length of the long time window is marked as nl; where m is the sampling time;

[0037] By the formula STA = 1 / ns × ∑ ns |x(m)| calculates the average value of the short-time window signal;

[0038] Through the formula LTA = 1 / nl × ∑ nl |x(m)| calculates the average value of the long-time window signal;

[0039] The trigger threshold is set to λ; through the formula Calculate the change value of the signal amplitude;

[0040] Determine whether the change value R1 of the signal amplitude is greater than or equal to λ; if yes, it is preliminarily determined that the pre-water hammer phenomenon occurs, and the pre-water hammer phenomenon is secondarily determined by the weighted coefficient method; if no, the pre-water hammer phenomenon does not occur.

[0041] The present invention provides a sliding long time window, and then takes a short time window within this window. The end points or starting points of the two windows coincide, and the ratio of the short time window signal average value (STA) to the long time window signal average value (LTA) is used to reflect the change in signal amplitude or energy; STA reflects the average value of the water hammer signal, and LTA reflects the average value of the background noise. At the time when the water hammer signal arrives, STA changes faster than LTA, and the corresponding STA / LTA value will have an obvious increase. When the ratio is greater than the set threshold, it can be determined that a water hammer event has occurred, thereby achieving the purpose of automatically detecting and picking up water hammer.

[0042] Preferably, the method for obtaining the characteristic function value includes:

[0043] The original time series is acquired through the pressure sensor and marked as Y(m);

[0044] The characteristic function value is calculated by the formula x(m)=|Y(m)-Y(m-1)|.

[0045] Preferably, the secondary judgment of the pre-water hammer phenomenon by weighted coefficient method includes:

[0046] By formula Calculate the standard deviation of the signal amplitude within the short time window;

[0047] By formula Calculate the standard deviation of the signal amplitude within the long time window;

[0048] By formula The ratio of the standard deviation of the signal amplitude in the long and short time windows is calculated;

[0049] By formula The change value of the amplitude of the secondary signal is calculated;

[0050] It is determined whether the change value R2 of the secondary signal amplitude is greater than or equal to λ; if yes, the secondary determination indicates that a pre-water hammer phenomenon occurs; if no, a pre-water hammer phenomenon does not occur.

[0051] The present invention measures the degree to which the signal amplitude deviates from the mean by calculating the standard deviation; moreover, when the water hammer signal arrives, the signal standard deviation in the short time window increases much more than the signal standard deviation in the long time window, and at this time a increases by a large margin, and at the same time a has an amplifying effect on the R2 value when the interference signal increases, but the amplifying effect on the R2 value when the water hammer signal arrives is more obvious, thereby greatly enhancing the difference in R value between the noise and the water hammer signal, and effectively improving the sensitivity of water hammer signal detection; the secondary judgment is conducive to further improving the water hammer recognition efficiency and reducing the missed judgment rate of water hammer event detection.

[0052] Preferably, the combining square correlation coefficient to determine whether the pre-water hammer phenomenon occurs includes:

[0053] A set of pressure data sequences that trigger water hammer and the current pressure data sequence to be measured are retrieved from the database and marked as Q and C respectively, where the length corresponding to sequence Q is n and the length corresponding to sequence Q is m;

[0054] By formula The Euclidean distance of the pressure data in the two sequences is calculated;

[0055] By obtaining the Euclidean distance, an n×m matrix grid is constructed. The first node in the matrix grid is set to d(q1, c1), and the next forward node is d(q i+1 , c j ), d(q i , c j+1 ), d(q i+1 , c j+1 ) and traverse all nodes in turn to obtain the shortest path;

[0056] The square correlation coefficient corresponding to the shortest path is obtained, and it is determined whether the square correlation coefficient is greater than or equal to 0.6; if yes, the pre-water hammer phenomenon occurs; if no, the pre-water hammer phenomenon does not occur.

[0057] The present invention calculates the Euclidean distance of the pressure data in the two sequences, which means aligning the pressure data in the two sequences; by finding the shortest path, the original pressure data can be processed into new pressure sequence data through the shortest path, thereby calculating the square correlation coefficient of the regularized pressure sequence data, and measuring the similarity between the two pressure data sequences through the square correlation coefficient, so as to quickly determine whether a water hammer event occurs based on the early warning signal.

[0058] Preferably, judging whether a water hammer event occurs according to the early warning signal includes:

[0059] Retrieve all types of warning signals and determine whether the number of warning signal types exceeds two or more; if yes, water hammer occurs; if no, water hammer does not occur; among them, the warning signal types are: warning signal one, warning signal two, and warning signal three.

[0060] The present invention uses a variety of methods to judge water hammer events, such as a steady-state pressure multiple method, a pressure slope method, a long and short time window method, and a dynamic time warping algorithm, which can verify each other and improve the accuracy and reliability of the judgment.

[0061] To achieve the above-mentioned purpose, the second aspect of the present invention provides a water hammer early warning system for a water supply network, comprising: a data acquisition module, a trigger mechanism determination module, and a water hammer phenomenon determination module;

[0062] Data acquisition module: instantaneous pressure data collected in real time by sensors;

[0063] Trigger mechanism determination module: Determine the water hammer trigger mechanism through the steady-state pressure multiple method, and determine whether the first water hammer state occurs based on the water hammer trigger mechanism; if yes, process it through the water hammer phenomenon determination module; if not, continue to determine;

[0064] Water hammer phenomenon judgment module: retrieve the water hammer signal that triggers the first water hammer state, and judge whether the second water hammer state occurs by the pressure slope method; if yes, the pre-water hammer phenomenon occurs, and the first warning signal is issued; if no, the pre-water hammer phenomenon does not occur;

[0065] Water hammer phenomenon is identified by long and short time window method; whether pre-water hammer phenomenon occurs is determined; if yes, the second warning signal is issued; if no, pre-water hammer phenomenon does not occur;

[0066] The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues;

[0067] Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] The present invention is based on data mining and statistical methods, and utilizes technologies such as the Internet of Things, cloud platforms, and big data to automatically update and generate a normal pressure fluctuation range and identify abnormal pressure fluctuation data in real time, thereby reducing the workload of management personnel and avoiding a large amount of waste of human and material resources; fully considering the propagation characteristics of water hammer waves, and timely and accurately capturing water hammer fluctuation characteristics; by using the mean of the median data in the median series as the steady-state pressure value, the influence of individual abnormal data on the calculation of the steady-state pressure value can be reduced, thereby improving the accuracy of monitoring; by judging the stability of the median series, it can ensure that trigger judgment is performed in a stable pressure environment, thereby avoiding false triggering due to pressure fluctuations.

[0070] The present invention adopts multiple methods such as steady-state pressure multiple method, pressure slope method, long and short time window method and dynamic time warping algorithm to judge water hammer events, which can verify each other and improve the accuracy and reliability of judgment; at the same time, this multiple judgment mechanism helps to reduce false alarms and missed alarms, and ensure the accuracy of early warning signals; the system can flexibly adjust the judgment criteria according to different judgment methods to adapt to the water hammer event monitoring needs under different pipeline systems and operating conditions; by calculating the Euclidean distance of the pressure data in the two sequences, it means aligning the pressure data in the two sequences; by finding the shortest path, the original pressure data is processed into new pressure sequence data through the shortest path, thereby calculating the square correlation coefficient of the corresponding regularized sequence signal, and measuring the similarity between the two pressure data sequences through the square correlation coefficient, so as to quickly judge whether a water hammer event occurs according to the early warning signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0072] Figure 1 It is a schematic diagram of the process of water hammer warning of the present invention;

[0073] Figure 2 A schematic diagram of the steps of identifying water hammer based on different methods under the first water hammer state according to the present invention;

[0074] Figure 3 It is a schematic diagram of water hammer simulation using the pressure slope method of the present invention;

[0075] Figure 4 It is a schematic diagram of water hammer simulation using the long and short time window method of the present invention;

[0076] Figure 5 It is a schematic diagram of identifying water hammer based on the shortest path of the present invention;

[0077] Figure 6 This is a schematic diagram of the relationship between the modules included in the present invention. DETAILED DESCRIPTION

[0078] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0079] See also Figure 1 The first embodiment of the present invention provides a water hammer early warning method for a water supply network, comprising:

[0080] The instantaneous pressure data collected by the sensor in real time is used to determine the water hammer triggering mechanism through the steady-state pressure multiple method;

[0081] Determine whether the first water hammer state occurs based on the water hammer trigger mechanism;

[0082] If yes, the pressure slope method is used to determine whether the second water hammer state has occurred; if yes, the pre-water hammer phenomenon is marked as occurring and a warning signal 1 is issued; if no, the pre-water hammer phenomenon is marked as not occurring;

[0083] If not, continue to judge;

[0084] Retrieve the water hammer signal that triggers the first water hammer state, and identify the water hammer phenomenon by using the long and short time window method; determine whether the pre-water hammer phenomenon occurs; if yes, issue the second warning signal; if no, mark the pre-water hammer phenomenon as not occurring;

[0085] The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues;

[0086] Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

[0087] See also Figure 2 , based on the steps of identifying water hammer by different methods under the first water hammer state, storing a number of data according to a set period to construct a pressure series, obtaining the median of the pressure data in the corresponding pressure series, and storing it in the pressure median series; obtaining the mean of the median data in the pressure median series, and taking the mean as the steady-state pressure value;

[0088] The trigger judgment mechanism is set as follows: the high pressure threshold is set to 1.06 times the steady-state pressure value, and the low pressure threshold is set to 0.94 times the steady-state pressure value; the pressure data in the pressure series is greater than or equal to the high pressure threshold or the pressure data is lower than the low pressure threshold;

[0089] Retrieve the data from the median series and label it as Z k , mark the minimum value in the data of the median sequence as Z min , mark the maximum value in the data of the median sequence as Z max ; Wherein, k is a positive integer;

[0090] By formula Y k =(Z k -Z min ) / (Z max-Z min ) Calculate the normalized standard deviation;

[0091] Determine whether the standard deviations are all less than 0.1; if so, determine in turn whether the pressure data in the pressure series reaches the trigger judgment mechanism; if so, the first water hammer state is triggered; if not, continue to judge;

[0092] If no, the first water hammer state is not triggered, the median sequence is updated and the judgment is made again;

[0093] Retrieve the pressure data corresponding to the first water hammer state trigger, and store the corresponding pressure data in the water hammer sequence until the pressure data in the water hammer sequence reaches the set length; set the set length as the water hammer detection window time length and mark it as hammer width ; and mark the pressure slope threshold as slope, obtain all pressure data within the set time after the trigger, mark the set time as t, and mark the maximum pressure value in all data as P max , the minimum pressure value is marked as P min ; Among them, hammer width >t>0;

[0094] By formula jump value =|P max -P min |Calculate the pressure difference;

[0095] By formula slope press =jump value / hammer width The real-time pressure slope is calculated;

[0096] The time corresponding to the maximum pressure value is marked as t1, and the time corresponding to the minimum pressure value is marked as t2;

[0097] Set the time difference threshold and calculate the time difference using the formula Δt=|t1-t2|;

[0098] Determine the slope press Is it greater than the slope? If yes, then determine whether Δt is greater than the time difference threshold; if yes, then mark the second water hammer state as occurring; if no, then mark the second water hammer state as not occurring;

[0099] If not, the second water hammer state is marked as not occurring;

[0100] After the trigger mechanism is implemented, the characteristic function value of the water hammer signal at time m is obtained and marked as x(m), and the characteristic function value is calculated by the formula x(m)=|Y(m)-Y(m-1)|; the length of the short time window is marked as ns, and the length of the long time window is marked as nl; where m is the sampling time;

[0101] By the formula STA = 1 / ns × ∑ ns |x(m)| calculates the average value of the short-time window signal;

[0102] Through the formula LTA = 1 / nl × ∑ nl |x(m)| calculates the average value of the long-time window signal;

[0103] The trigger threshold is set to λ; through the formula Calculate the change value of the signal amplitude;

[0104] Determine whether the change value R1 of the signal amplitude is greater than or equal to λ; if yes, it is preliminarily determined that the pre-water hammer phenomenon occurs, and the pre-water hammer phenomenon is secondarily determined by the weighted coefficient method; if no, the pre-water hammer phenomenon does not occur;

[0105] By formula Calculate the standard deviation of the signal amplitude within the short time window;

[0106] By formula Calculate the standard deviation of the signal amplitude within the long time window;

[0107] By formula The ratio of the standard deviation of the signal amplitude in the long and short time windows is calculated;

[0108] By formula The change value of the amplitude of the secondary signal is calculated;

[0109] It is determined whether the change value R2 of the secondary signal amplitude is greater than or equal to λ; if yes, the secondary judgment concludes that a pre-water hammer phenomenon has occurred, and a second warning signal is issued; if no, no pre-water hammer phenomenon has occurred.

[0110] For example, in the existing scheme condition 1, the design is to simulate the water hammer of the rapid closing valve on the DN800 valve of the water supply main system. The valve is located about 150m away from the water plant. The change of the water hammer pressure in the pipeline network is related to the valve closing time and law. It is assumed that the valve opening is linearly related to the flow rate. The specific valve operation mode is as follows: the throttle valve is quickly closed to 80% in 5s, and then slowly closed to 100% within 8s. The initial valve opening is set to 100% to simulate the valve closing water hammer.

[0111] Through the historical data of 30 seconds before the current moment, the median of the 25Hz sampling data per second is dynamically calculated and the mean is calculated to obtain 0.12Mpa, and 0.12Mpa is used as the steady-state pressure value at this moment; the trigger judgment mechanism is set as follows: the high pressure threshold is set to 1.06 times the steady-state pressure value, that is, 0.1272, and the low pressure threshold is set to 0.94 times the steady-state pressure value, that is, 0.1128; the pressure data in the pressure series is greater than or equal to the high pressure threshold or the pressure data is lower than the low pressure threshold;

[0112] When the instantaneous pressure triggers the high or low pressure threshold, obtain 50 data within 2-3 seconds after the trigger; retrieve the data in the median series and mark it as Z k , mark the minimum value in the data of the median sequence as Z min , mark the maximum value in the data of the median sequence as Z max ; Wherein, k is a positive integer;

[0113] By formula Y k =(Z k -Z min ) / (Z max -Z min ) The normalized standard deviation values ​​calculated include 20 values ​​of 0.08, 15 values ​​of 0.05, 8 values ​​of 0.09 and 7 values ​​of 0.01; all standard deviations are less than 0.1;

[0114] At the same time, if there is a pressure value greater than or equal to 0.1272Mpa in the pressure series, the first water hammer state is triggered;

[0115] Retrieve the pressure data corresponding to the first water hammer state trigger, and store the corresponding pressure data in the water hammer sequence until the pressure data in the water hammer sequence reaches the set length; set the set length as the water hammer detection window time length and mark it as hammer width ; Default hammer width =2s, preset slope = 0.05MPa / s

[0116] Get all pressure data within 1s after triggering (such as Figure 3 ),Depend on Figure 3 Get the maximum pressure value and the minimum pressure value in all data;

[0117] By formula jump value =|P max -P min |The calculated pressure difference is 151.0Mpa;

[0118] By formula slope press =jump value / hammer width The calculated real-time pressure slope is 75.5Mpa / s;

[0119] The time corresponding to the maximum pressure value is marked as t1 = 0.1s, and the time corresponding to the minimum pressure value is marked as 0.25s;

[0120] The time difference threshold is set to 0.08s, and the time difference is calculated to be 0.15s using the formula Δt=|t1-t2|;

[0121] If the real-time pressure slope is 75.5 MPa / s greater than 0.05 MPa / s, and 0.15s greater than 0.08s, pre-water hammer occurs and an early warning signal 1 is issued;

[0122] Obtained through third-party software based on long and short time window method Figure 4 , the trigger threshold is set to 2.0, by Figure 4 It can be obtained that when R2=10.9 is greater than 2.0, pre-water hammer phenomenon occurs and the second warning signal is issued.

[0123] In the present invention, a high pressure threshold and a low pressure threshold are set as a trigger judgment mechanism, which can monitor in real time whether the pressure data exceeds the normal range, so as to respond to the water hammer state in time; by using the median and the mean to calculate the steady-state pressure value, and setting a reasonable trigger threshold, the possibility of false alarm and missed alarm is reduced; at the same time, judging the stability of the median series can further reduce the false triggering caused by pressure fluctuations.

[0124] See also Figure 5 , based on the shortest path to identify water hammer, a set of pressure data sequences that trigger water hammer and the current pressure data sequence to be tested are retrieved from the database, marked as Q and C respectively, where the length corresponding to sequence Q is n and the length corresponding to sequence Q is m;

[0125] By formula The Euclidean distance of the pressure data in the two sequences is calculated;

[0126] By obtaining the Euclidean distance, an n×m matrix grid table is constructed (as shown in Table 1).

[0127] Table 1 Matrix grid table

[0128] d(q1,c1) d(q1,c2) … d(q1, cm) d(q2,c1) d(q2,c2) … d(q2, cm) … … … … d(qn,c1) d(qn,c2) … d(qn,cm)

[0129] Assume that the first node in the matrix grid is d(q1, c1), then the next forward node is d(q i+1 , c j ), d(q i , c j+1 ), d(q i+1 , c j+1 ) and traverse all nodes in turn to obtain the shortest path;

[0130] The square correlation coefficient corresponding to the shortest path is obtained, and it is determined whether the square correlation coefficient is greater than or equal to 0.6; if yes, the pre-water hammer phenomenon occurs; if no, the pre-water hammer phenomenon does not occur.

[0131] For example, a set of pressure data sequences that trigger water hammer and the current pressure data sequence to be measured are retrieved from the database, marked as Q and C respectively, where the length corresponding to sequence Q is n and the length corresponding to sequence Q is m; by obtaining the Euclidean distance, an n×m matrix grid table is constructed (such as Figure 3 As shown); obtain the first node d(q1, c1) = 2, d(2, 1) = 1, d(1, 2) = 3, d(2, 2) = 4, according to the shortest path principle, select d(2, 1) as the second node, at this time d(3, 1) = 5, d(1, 3) = 4, d(3, 3) = 8, select d(1, 3) as the third node, traverse all the data in the table in turn, and thus obtain the shortest path; the original pressure data can be processed into new pressure sequence data through the shortest path, so as to obtain the square correlation coefficient corresponding to the regularized sequence signal of 0.59 is less than 0.6, then no pre-water hammer phenomenon occurs.

[0132] See also Figure 6 , the second aspect of the present invention provides a water hammer early warning system for a water supply network, including: a data acquisition module, a trigger mechanism determination module, and a water hammer phenomenon determination module;

[0133] Data acquisition module: instantaneous pressure data collected in real time by sensors;

[0134] Trigger mechanism determination module: Determine the water hammer trigger mechanism through the steady-state pressure multiple method, and determine whether the first water hammer state occurs based on the water hammer trigger mechanism; if yes, process it through the water hammer phenomenon determination module; if not, continue to determine;

[0135] Water hammer phenomenon judgment module: retrieve the water hammer signal that triggers the first water hammer state, and judge whether the second water hammer state occurs by the pressure slope method; if yes, the pre-water hammer phenomenon occurs, and the first warning signal is issued; if no, the pre-water hammer phenomenon does not occur;

[0136] Water hammer phenomenon is identified by long and short time window method; whether pre-water hammer phenomenon occurs is determined; if yes, the second warning signal is issued; if no, pre-water hammer phenomenon does not occur;

[0137] The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues;

[0138] Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

[0139] For example, the warning signal type is retrieved, and if there are two types of warning signals, warning signals one and two, a water hammer event occurs at the corresponding moment, and a warning signal is immediately sent to the staff.

[0140] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0141] The working principle of the present invention is as follows: instantaneous pressure data collected in real time by the sensor; the water hammer triggering mechanism is determined by the steady-state pressure multiple method, and whether the first water hammer state occurs is judged based on the water hammer triggering mechanism; if yes, the water hammer phenomenon judgment module is entered; if no, the judgment is continued; the water hammer signal that triggers the first water hammer state is retrieved, and whether the second water hammer state occurs is judged by the pressure slope method; if yes, the pre-water hammer phenomenon occurs, and the first warning signal is issued; if no, the pre-water hammer phenomenon does not occur;

[0142] Water hammer phenomenon is identified by long and short time window method; whether pre-water hammer phenomenon occurs is determined; if yes, the second warning signal is issued; if no, pre-water hammer phenomenon does not occur;

[0143] The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues;

[0144] Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

[0145] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A water hammer early warning method for a water supply network, characterized in that: include: The instantaneous pressure data collected by the sensor in real time is used to determine the water hammer triggering mechanism through the steady-state pressure multiple method; Determine whether the first water hammer state occurs based on the water hammer trigger mechanism; Yes, use the pressure slope method to determine whether the second water hammer state occurs; If yes, the pre-water hammer phenomenon is marked as occurring and a warning signal 1 is issued; if no, the pre-water hammer phenomenon is marked as not occurring; If not, continue to judge; Retrieve the water hammer signal that triggers the first water hammer state, and identify the water hammer phenomenon by using the long and short time window method; determine whether the pre-water hammer phenomenon occurs; if yes, issue the second warning signal; if no, mark the pre-water hammer phenomenon as not occurring; The shortest path is obtained through the dynamic time warping algorithm, and the regularized pressure data is obtained through the shortest path. Combined with the square correlation coefficient, it is judged whether the pre-water hammer phenomenon occurs; if yes, the third warning signal is issued; if not, the judgment continues; Determine whether a water hammer event has occurred based on the warning signal; if yes, send the warning signal to the staff; if no, continue to judge.

2. A water hammer early warning method for a water supply network according to claim 1, characterized in that: The determining whether the first water hammer state occurs based on the water hammer trigger mechanism includes: Store a number of data according to a set period to construct a pressure series, obtain the median of the pressure data in the corresponding pressure series, and store it in the pressure median series; obtain the mean of the median data in the pressure median series, and use the mean as the steady-state pressure value; The trigger judgment mechanism is set as follows: the high pressure threshold is set to 1.06 times the steady-state pressure value, and the low pressure threshold is set to 0.94 times the steady-state pressure value; the pressure data in the pressure series is greater than or equal to the high pressure threshold or the pressure data is lower than the low pressure threshold; Determine whether the median sequence is stable; if so, determine in turn whether the pressure data in the pressure sequence reaches the trigger judgment mechanism; if so, mark the first water hammer state as triggered; if not, continue to judge; If not, the first water hammer state is marked as not triggered, the median sequence is updated and the judgment is made again.

3. A water hammer early warning method for a water supply network according to claim 2, characterized in that: The determining whether the median sequence is stable includes: Retrieve the data from the median series and label it as Z k , mark the minimum value in the data of the median sequence as Z min , mark the maximum value in the data of the median sequence as Z max ; Wherein, k is a positive integer; By formula Y k =(Z k -Z min ) / (Z max -Z min ) Calculate the normalized standard deviation; Determine whether the standard deviations are all less than 0.1; if yes, the median sequence is stable; if no, the median sequence is unstable.

4. A water hammer early warning method for a water supply network according to claim 1, characterized in that: The determining whether the second water hammer state occurs by the pressure slope method includes: Retrieve the pressure data corresponding to the first water hammer state trigger, and store the corresponding pressure data in the water hammer sequence until the pressure data in the water hammer sequence reaches the set length; set the set length as the water hammer detection window time length and mark it as hammer width ; and mark the pressure slope threshold as slope, obtain all pressure data within the set time after the trigger, mark the set time as t, and mark the maximum pressure value in all data as P max , the minimum pressure value is marked as P min ; Among them, hammer width >t>0; By formula jump value =|P max -P min |Calculate the pressure difference; By formula slope press =jump value / hammer width The real-time pressure slope is calculated; The time corresponding to the maximum pressure value is marked as t1, and the time corresponding to the minimum pressure value is marked as t2; Set the time difference threshold and calculate the time difference using the formula Δt=|t1-t2|; Determine the slope press Is it greater than the slope? If yes, it is determined whether Δt is greater than the time difference threshold; if yes, the second water hammer state is marked as occurring; if no, the second water hammer state is marked as not occurring; If not, the second water hammer state is marked as not occurring.

5. A water hammer early warning method for a water supply network according to claim 1, characterized in that: The water hammer phenomenon identification by long and short time window method includes: After the trigger mechanism is implemented, the characteristic function value of the water hammer signal at time m is obtained and marked as x(m); the length of the short time window is marked as ns, and the length of the long time window is marked as nl; where m is the sampling time; By the formula STA = 1 / ns × ∑ ns |x(m)| calculates the average value of the short-time window signal; Through the formula LTA = 1 / nl × ∑ nl |x(m)| calculates the average value of the long-time window signal; The trigger threshold is set to λ; through the formula Calculate the change value of the signal amplitude; Determine whether the change value R1 of the signal amplitude is greater than or equal to λ; if yes, it is preliminarily determined that the pre-water hammer phenomenon occurs, and the pre-water hammer phenomenon is secondarily determined by the weighted coefficient method; if no, the pre-water hammer phenomenon does not occur.

6. A water hammer early warning method for a water supply network according to claim 5, characterized in that: The method for obtaining the characteristic function value includes: The original time series is acquired through the pressure sensor and marked as Y(m); The characteristic function value is calculated by the formula x(m)=|Y(m)-Y(m-1)|.

7. A water hammer early warning method for a water supply network according to claim 5, characterized in that: The secondary judgment of the pre-water hammer phenomenon by the weighted coefficient method includes: By formula Calculate the standard deviation of the signal amplitude within the short time window; By formula Calculate the standard deviation of the signal amplitude within the long time window; By formula The ratio of the standard deviation of the signal amplitude in the long and short time windows is calculated; By formula The change value of the amplitude of the secondary signal is calculated; It is determined whether the change value R2 of the secondary signal amplitude is greater than or equal to λ; if yes, the secondary determination indicates that a pre-water hammer phenomenon occurs; if no, a pre-water hammer phenomenon does not occur.

8. A water hammer early warning method for a water supply network according to claim 1, characterized in that: The combined square correlation coefficient is used to determine whether the pre-water hammer phenomenon occurs, including: A set of pressure data sequences that trigger water hammer and the current pressure data sequence to be measured are retrieved from the database and marked as Q and C respectively, where the length corresponding to sequence Q is n and the length corresponding to sequence Q is m; By formula The Euclidean distance of the pressure data in the two sequences is calculated; By obtaining the Euclidean distance, an n×m matrix grid is constructed. The first node in the matrix grid is set to d(q1, c1), and the next forward node is d(q i+1 , c j ), d(q i , c j+1 ), d(q i+1 , c j+1 ) and traverse all nodes in turn to obtain the shortest path; The square correlation coefficient corresponding to the shortest path is obtained, and it is determined whether the square correlation coefficient is greater than or equal to 0.6; if yes, the pre-water hammer phenomenon occurs; if no, the pre-water hammer phenomenon does not occur.

9. A water hammer early warning method for a water supply network according to claim 1, characterized in that: The determining whether a water hammer event occurs according to the early warning signal includes: Retrieve all types of warning signals and determine whether the number of warning signal types exceeds two or more; if yes, water hammer occurs; if no, water hammer does not occur; among them, the warning signal types are: warning signal one, warning signal two, and warning signal three.

10. A water hammer early warning system for a water supply network, applied to a water hammer early warning method for a water supply network as claimed in any one of claims 1 to 9, characterized in that: include: Data acquisition module, trigger mechanism determination module, water hammer phenomenon determination module; Data acquisition module: instantaneous pressure data collected in real time by sensors; Trigger mechanism determination module: determines the water hammer trigger mechanism through the steady-state pressure multiple method, and determines whether the first water hammer state occurs based on the water hammer trigger mechanism; Water hammer phenomenon judgment module: retrieve the water hammer signal that triggers the first water hammer state, and judge whether the second water hammer state occurs by the pressure slope method; Identify water hammer phenomenon through long and short time window method; judge whether pre-water hammer phenomenon occurs; The shortest path is obtained by dynamic time warping algorithm, and the normalized pressure data is obtained by the shortest path, and the square correlation coefficient is combined to determine whether the pre-water hammer phenomenon occurs; and Determine whether a water hammer event occurs based on the early warning signal.

Images