Comprehensive inverse time limit alarm method for multi-target monitoring data of water supply network

Abstract

A comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network comprises the following steps: step 1, collecting data such as node flow, water pressure, unit state, pressure drop duration and the like of the water supply network, and respectively marking the data as normal and abnormal according to actual conditions; 2, inverse time limit formulas of water flow, water pressure, unit temperature and the like are set, and formula parameters are automatically set by using a seagull algorithm according to pipe network data; 3, receiving various monitoring data by the system, and calculating alarm delay time Ti according to respective inverse time limit formulas to obtain a comprehensive time limit T; 4, when the data exceed the limit, starting a timer Timer; if the data is recovered to be normal within the timing time, the timer and the alarm delay time are reset; and otherwise, if the timing time Timer is greater than the alarm delay time T, the system gives an alarm to remind the staff to go for troubleshooting.

Description

Technical Field

[0001] The invention relates to the technical field of gas overhead pipelines, in particular to a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network. Background Art

[0002] As people's living standards gradually improve, residents have higher requirements for the quality of water supply services. Stable water pressure in the water supply network is one of the keys to ensuring water supply quality. Most water supply plants now use a constant pressure variable frequency water supply system. The principle of this system is to use the water pressure feedback closed-loop control of the water supply pipeline to maintain the stability of the pipeline water pressure to meet the user's water demand. In the case of a long water supply network and a wide radiation range, most smart water systems use wireless remote transmission to achieve long-distance real-time pressure monitoring, and issue an alarm message when the water supply pressure value is abnormal.

[0003] The common water pressure abnormality alarm strategy is to set certain upper and lower limits of water pressure. When the water pressure monitoring data is lower than the lower limit or exceeds the upper limit, an alarm is issued to remind the staff to check the abnormality. However, the actual water pressure changes are complex and it is very easy to trigger the water pressure abnormality alarm by mistake. For example, a surge in residents' water consumption at a certain point in time or multiple sprinkler trucks receiving water will cause the water pressure at the pipeline node to drop for a short time, and the abnormality alarm will be triggered when it is lower than the lower limit of the water pressure. The water pressure curve of the secondary booster pump station of the water supply pipeline network of the Gezhouba Hub Water Plant is shown in Figure 1 As can be seen from the figure, the water pressure is below the lower limit for a short period of time many times in a day. This situation does not affect the normal water supply, but it will trigger abnormal alarms many times. If the lower limit is set to reduce false triggers, the sensitivity and reliability of the abnormal water pressure alarm will be lost, and the water pressure quality cannot be guaranteed to be stable. It can be seen that the selectivity, reliability and sensitivity of a single water pressure upper and lower limit alarm strategy are all low. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, improve the accuracy, reliability and adaptability of abnormal alarms of water supply network data, provide effective decision support for water workers, shorten fault discovery and repair time, and reduce water resource waste and economic losses.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a water supply network multi-target monitoring data comprehensive inverse time limit alarm method, comprising the following steps: Step 1: Collect data such as flow rate, water pressure, unit status, and pressure drop duration at water supply network nodes, and mark them as “normal” or “abnormal” according to actual conditions; Step 2: Set the inverse time formulas for water flow, water pressure, unit temperature, etc., and use the Seagull algorithm to automatically set the formula parameters based on the pipe network data; Step 3: The system receives various monitoring data, calculates the alarm delay time Ti according to the respective inverse time limit formulas, and obtains the comprehensive time limit T; Step 4: When the data exceeds the limit, start the timer Timer; if the data returns to normal within the timer, the timer and alarm delay time are reset; otherwise, if the timer Timer> alarm delay time T, the system issues an alarm to remind the staff to go for investigation.

[0006] Preferably, in step 2, the inverse time formula of pipeline water flow is: ; T 水流量 is the water flow delay, Q is the water flow, Qp is the water flow setting value, and Tp is the time adjustment constant; The flow data of several network nodes are classified, and +1 is marked as "normal" and -1 is marked as "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit flow is Qx, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ; In summary, we can get: ; According to the classification formula, the classification objective function is set as: ; Assume the adjustment parameter Tp as a variable, use the Seagull algorithm to optimize and obtain the optimal parameter of Tp.

[0007] Preferably, in step 2, the pipeline pressure inverse time formula is: ; Where T 水压 is the water pressure delay time (min), Tmax is the maximum delay time (min); α is the change adjustment coefficient; P is the real-time pressure (Mpa); Pset is the pressure threshold (Mpa); The water pressure sudden change data of several pipe network nodes are classified, marked as +1 for "normal" and -1 for "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit pressure is Px, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ; In summary, we can get: ; According to the classification formula, the classification objective function is set as: ; Assume adjustment parameters Tmax and α as variables, use Seagull algorithm for optimization, and obtain the optimal parameters of Tmax and α.

[0008] The inverse time formula for unit temperature and parameter setting method are the same as the inverse time formula for water pressure.

[0009] Preferably, the Seagull algorithm includes migration and attack behaviors: 1. Migration Behavior In order to avoid collision between each seagull and the surrounding seagulls, the position of the seagull is adjusted through variable A; ; Where: CS(t) is the new position that does not collide with other seagulls; PS(t) is the current position of the seagull; A is the movement behavior of the seagull; t is the current iteration number, and the size of A is controlled by fc; ; Among them: Kmax is the maximum number of iterations; fc is 2. Then all seagulls are moved closer to the best seagull position.

[0010] ; ; Where: Ms(t) is the direction of the best seagull; Ps(t) is the position of the best seagull; B is an important random parameter responsible for balancing the global and local searches of the algorithm, and rand is a random number in the range of [0,1]; the seagull moves towards the position of the best seagull and reaches the new position Ds(t); ; 2. Aggressive Behavior When attacking prey, the seagulls move in a spiral shape in the air. The attack positions of the seagulls are: ; Where: r is the spiral radius; k is a random angle value between 0 and 2π; u and v are related constants of the spiral shape; e is the base of the natural logarithm.

[0011] Preferably, in step 3, the comprehensive time limit T is calculated ; Where n is the number of monitored data types, Calculate the results of the inverse time formula for each monitoring data.

[0012] The present invention provides a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, which has the following beneficial effects: 1. Avoids false triggering of the water supply network alarm system and saves system resources; 2. Improve the accuracy and reliability of abnormal alarms for water supply network data, and provide effective decision-making support for staff; 3. Shorten the repair time of water supply network failure, reduce water resource waste and economic losses; 4. Use the Seagull algorithm to automatically set parameters and reduce manual calculation costs; 5. Set the alarm time according to the actual water supply data to improve the accuracy of judgment; 6. Comprehensively judge abnormalities by integrating various pipe network data, thus improving the system's preventive capability and adaptability; 7. The comprehensive time limit formula reduces the system calculation cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 It is the water pressure curve diagram of the existing water supply network; Figure 2 It is the inverse time characteristic curve diagram of water supply flow rate of the present invention; Figure 3 It is a delay time-water pressure curve diagram under different variable adjustment coefficients of the present invention; Figure 4 This is a graph of the alarm delay time-water pressure curve of the present invention; Figure 5 This is a structural diagram of the comprehensive inverse time limit strategy of the present invention; Figure 6 It is the flow inverse time curve diagram of the present invention; Figure 7 It is the water pressure inverse time curve diagram of the present invention; Figure 8a This is a diagram showing the pressure variation of the water pipe network of the sprinkler truck of the present invention; Figure 8b This is the alarm delay diagram of the water pipe network for the sprinkler truck of the present invention; Figure 9a This is a pressure change diagram of the booster pump station unit failure shutdown of the present invention; Figure 9b This is the booster pump station unit fault shutdown alarm delay diagram of the present invention. DETAILED DESCRIPTION

[0014] A method for comprehensive inverse time limit alarm of multi-target monitoring data of a water supply network comprises the following steps: Step 1: Collect data such as flow rate, water pressure, unit status, and pressure drop duration at water supply network nodes, and mark them as “normal” or “abnormal” according to actual conditions; Step 2: Set the inverse time formulas for water flow, water pressure, unit temperature, etc., and use the Seagull algorithm to automatically set the formula parameters based on the pipe network data; Step 3: The system receives various monitoring data, calculates the alarm delay time Ti according to the respective inverse time limit formulas, and obtains the comprehensive time limit T; Step 4: When the data exceeds the limit, start the timer Timer; if the data returns to normal within the timer, the timer and alarm delay time are reset; otherwise, if the timer Timer> alarm delay time T, the system issues an alarm to remind the staff to go for investigation.

[0015] Preferably, in step 2, the inverse time formula of pipeline water flow is: ; T 水流量 is the water flow delay, Q is the water flow, Qp is the water flow setting value, and Tp is the time adjustment constant; Set the water supply flow setting value Qp to 500m 3* h -1 The inverse time characteristic curves when Tp is 0.01, 0.1 and 1 are as follows: Figure 2 As shown in the figure, the smaller the Tp setting, the faster the delay time changes, and according to the inverse time characteristic of the above formula, when the water flow reaches 2*Qp (m 3* h -1 ), the delay time is 10*Tp (min), so Tp can be set according to the water flow rate and actual response time.

[0016] The flow data of several network nodes are classified, and +1 is marked as "normal" and -1 is marked as "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit flow is Qx, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ; In summary, we can get: ; According to the classification formula, the classification objective function is set as: ; Assume the adjustment parameter Tp as a variable, use the Seagull algorithm to optimize and obtain the optimal parameter of Tp.

[0017] Preferably, in step 2, the pipeline pressure inverse time formula is: ; Where T 水压is the water pressure delay time (min), Tmax is the maximum delay time (min); α is the change adjustment coefficient; P is the real-time pressure (Mpa); Pset is the pressure threshold (Mpa); In the above formula, the maximum delay time Tmax and the change adjustment coefficient α need to be set according to the actual water usage changes. Taking the local water usage as an example, the main reasons for the sudden change of water pressure at the network node are shown in Table 1. The change adjustment coefficient α determines the rate at which the delay time changes with pressure. The larger the change adjustment coefficient, the faster the delay time decreases. Figure 3 After the parameters are set, the curve of alarm delay time and pipeline node pressure is as follows: Figure 4 shown.

[0018] Table 1 Main reasons for sudden changes in water pressure at pipe network nodes

[0019] The water pressure sudden change data of several pipe network nodes are classified, marked as +1 for "normal" and -1 for "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit pressure is Px, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ; In summary, we can get: ; According to the classification formula, the classification objective function is set as: ; Assume adjustment parameters Tmax and α as variables, use Seagull algorithm for optimization, and obtain the optimal parameters of Tmax and α.

[0020] The inverse time formula for unit temperature and parameter setting method are the same as the inverse time formula for water pressure.

[0021] Preferably, the Seagull algorithm includes migration and attack behaviors: 1. Migration Behavior In order to avoid collision between each seagull and the surrounding seagulls, the position of the seagull is adjusted through variable A; ; Where: CS(t) is the new position that does not collide with other seagulls; PS(t) is the current position of the seagull; A is the movement behavior of the seagull; t is the current iteration number, and the size of A is controlled by fc; ; Among them: Kmax is the maximum number of iterations; fc is 2. Then all seagulls are moved closer to the best seagull position.

[0022] ; ; Where: Ms(t) is the direction of the best seagull; Ps(t) is the position of the best seagull; B is an important random parameter responsible for balancing the global and local searches of the algorithm, and rand is a random number in the range of [0,1]; the seagull moves towards the position of the best seagull and reaches the new position Ds(t); ; 2. Aggressive Behavior When attacking prey, the seagulls move in a spiral shape in the air. The attack positions of the seagulls are: ; Where: r is the spiral radius; k is a random angle value between 0 and 2π; u and v are related constants of the spiral shape; e is the base of the natural logarithm.

[0023] Preferably, in step 3, the comprehensive time limit T is calculated ; Where n is the number of monitored data types, The calculation result of the inverse time formula for each monitoring data. From the above formula, it is easy to get: ; The structure of the comprehensive inverse time strategy is as follows Figure 5 As shown, in addition to flow rate over-limit and water pressure over-limit, other data such as unit temperature over-limit can also be monitored.

[0024] like Figure 6 As shown, according to the water flow data collected in step one, the Seagull algorithm is used for optimization to obtain the optimal parameter Tp=1.2, and the curve is shown in the figure.

[0025] like Figure 7 As shown, according to the water pressure data collected in step one, the Seagull algorithm is used for optimization to obtain the optimal parameters Tmax=40, α=1.44, and the curve is shown in the figure.

[0026] In order to verify the selectivity and reliability of the present invention, verification was carried out under two working conditions: a sprinkler truck receiving water and a booster pump station unit shutting down due to a fault. The system's data sampling and calculation frequency was 1 min / per unit. When the water pressure exceeded the limit, the system started delay time calculation and timing. When the timing time was greater than the delay time, the system sent an alarm message to the staff.

[0027] The pressure change and alarm delay of the water pipe network of the sprinkler truck, such as Figure 8a , 8b The water pressure curve is relatively stable as a whole, and the time below the lower limit is within 10 minutes. When the water pressure rises, the delay time also starts to rise. Before the water pressure returns to the lower limit, the timing curve and the delay time do not intersect, so the system does not issue an alarm; the pressure change and alarm delay of the secondary booster pump station unit failure shutdown of the water supply network, such as Figure 9a , 9b The water pressure curve dropped significantly, and during the recovery process, the water pressure fluctuated greatly due to the air in the pipeline. The time below the lower limit exceeded 40 minutes. When the timing curve intersected with the alarm delay curve at 4 minutes, the system immediately issued an alarm. After receiving the alarm information, the staff went to the site in time to repair the faulty unit, and the water pressure finally returned to normal.

[0028] In summary, in the inverse time limit alarm strategy, the water pressure fluctuation caused by normal water use, the water pressure exceeding the limit time will not exceed the alarm delay time, and the system will not issue an alarm; while the water pressure abnormality caused by the fault, the water pressure exceeding the limit value is large, and the exceeding limit time is much longer than the alarm delay time, and the system can quickly issue an alarm.

[0029] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A comprehensive inverse time limit alarm method for multi-target monitoring data of water supply network, It is characterized in that The following steps are involved: Step 1: Collect data on water supply network node flow, water pressure, unit status, and pressure drop duration, and mark them as "normal" or "abnormal" according to actual conditions; Step 2: Set the inverse time formula of water flow, water pressure and unit temperature, and use the Seagull algorithm to automatically set the formula parameters according to the pipe network data; Step 3: The system receives various monitoring data, calculates the alarm delay time Ti according to the respective inverse time limit formulas, and obtains the comprehensive time limit T; Step 4: When the data exceeds the limit, start the timer Timer; if the data returns to normal within the timer, the timer and alarm delay time are reset; otherwise, if the timer Timer> alarm delay time T, the system issues an alarm to remind the staff to go for investigation.

2. According to claim 1, a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, It is characterized in that In step 2, the inverse time formula of pipeline water flow is: ； T 水流量 is the water flow delay, Q is the water flow, Qp is the water flow setting value, and Tp is the time adjustment constant; The flow data of several network nodes are classified, and +1 is marked as "normal" and -1 is marked as "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit flow is Qx, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ； In summary, we can get: ； According to the classification formula, the classification objective function is set as: ； Assume the adjustment parameter Tp as a variable, use the Seagull algorithm to optimize and obtain the optimal parameter of Tp.

3. According to claim 1, a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, It is characterized in that In step 2, the inverse time formula of pipeline pressure is: ； Where T 水压 is the water pressure delay time, Tmax is the maximum delay time; α is the change adjustment coefficient; P is the real-time pressure; Pset is the pressure threshold; The water pressure sudden change data of several pipe network nodes are classified, marked as +1 for "normal" and -1 for "abnormal". Assume that there are n data samples in total, the maximum duration is Tx, the maximum over-limit pressure is Px, and the classification mark is y. For "normal" data, the delay time should be greater than its maximum duration, and for "abnormal" data, the delay time should be less than its maximum duration. The classification formula is obtained according to the above formula: ； In summary, we can get: ； According to the classification formula, the classification objective function is set as: ； Assume adjustment parameters Tmax and α as variables, use Seagull algorithm for optimization, and obtain the optimal parameters of Tmax and α.

4. According to claim 2 or 3, a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, It is characterized in that The Seagull algorithm includes migration and attack behaviors:

1. Migration Behavior In order to avoid collision between each seagull and the surrounding seagulls, the position of the seagull is adjusted through variable A; ； Where: CS(t) is the new position that does not collide with other seagulls; PS(t) is the current position of the seagull; A is the movement behavior of the seagull; t is the current iteration number, and the size of A is controlled by fc; ； Among them: Kmax is the maximum number of iterations; fc is 2. Then all seagulls are moved closer to the best seagull position. ； ； Where: Ms(t) is the direction of the best seagull; Ps(t) is the position of the best seagull; B is an important random parameter responsible for balancing the global and local searches of the algorithm, and rand is a random number in the range of [0,1]; the seagull moves towards the position of the best seagull and reaches the new position Ds(t); ； 2. Aggressive Behavior When attacking prey, the seagulls move in a spiral shape in the air. The attack positions of the seagulls are: ； Where: r is the spiral radius; k is a random angle value between 0 and 2π; u and v are related constants of the spiral shape; e is the base of the natural logarithm.

5. According to claim 1, a comprehensive inverse time limit alarm method for multi-target monitoring data of a water supply network, It is characterized in that In step 3, the comprehensive time limit T is calculated. ； Where n is the number of monitored data types, Calculate the results of the inverse time formula for each monitoring data.

Images