Methods and systems for demonstrating the performance of flow measurement equipment using the flow balance method

Abstract

This invention relates to a method and system for verifying the performance of flow measurement equipment using the flow balance method, comprising the following steps: Step S1: Establishing a loss model using historical data to verify the measurement results of the equipment itself. If the data mutation rate is acceptable, continue execution; otherwise, determine that the equipment is unreliable and terminate execution; Step S2: Analyzing the equipment; First, based on the loss model, calculate the logical value of the flow loss value using an AI algorithm; Then, compare the logical value with the downstream measured value to determine whether the equipment is reliable; If the downstream measured value is within the logical value range corresponding to the logical value, the equipment is determined to be reliable and execution continues; If the downstream measured value is not within the logical value range, the equipment is unreliable and execution terminates; Step S3: Extracting data from upstream and downstream stations, and performing cross-validation using the comparison method and threshold method to determine the reliability of the equipment; If the equipment is determined to be reliable, continue execution.

Description

Technical Field

[0001] This invention relates to a method and system for demonstrating the performance of flow measurement equipment using the flow balance method. Background Technology

[0002] In modern water management, enclosed artificial canals are widely used due to their ability to effectively reduce water evaporation, prevent pollution, and improve water conveyance efficiency. However, how to achieve accurate flow measurement under complex and variable environmental conditions to ensure an accurate and objective reflection of the on-site flow situation has become an urgent problem to be solved. Furthermore, because large flow measurements are involved, there is no accurate calibration result. Under such circumstances, ensuring the reliability of the measurement equipment's results is a pressing technical issue that needs to be addressed.

[0003] Existing flow measurement equipment generally determines its reliability based on the equipment's condition and related technical specifications. However, judging the reliability of an equipment solely based on its condition, without standard data for verification, makes it impossible to demonstrate the performance of the flow measurement equipment. Summary of the Invention

[0004] In general, the technical problem to be solved by this invention is to provide a method and system for verifying the performance of flow measurement equipment using the flow balance method. This method can reversely verify the reliability of the flow measurement data by using the measurement results of the flow measurement equipment without the need for third-party testing methods or measurement point calibration.

[0005] To solve the above problems, the technical solution adopted by the present invention is as follows: A method for demonstrating the performance of a flow measurement device using the flow balance method includes the following steps; Step S1: Build a loss model using historical data and verify the measurement results of the equipment itself. If the data mutation rate is acceptable, continue execution; otherwise, determine that the equipment is unreliable and end execution. Step S2: Analyze the equipment; First, based on the loss model, calculate the logical value S2_L of the traffic loss using an AI algorithm; Then, compare the logical value S2_L with the downstream measured value to determine whether the equipment is reliable; If the downstream measured value is within the logical value range corresponding to the logical value S2_L, the device is considered reliable and execution continues; if the downstream measured value is not within the logical value range, the device is considered unreliable and execution ends. Step S3: Extract data from upstream and downstream monitoring stations, and perform cross-validation to determine the reliability of the equipment using the comparison method and threshold method; if the equipment is determined to be reliable, continue execution; otherwise, end execution.

[0006] Furthermore, the method is applied to artificial canals; In step S1, first, the change process of water level or flow rate shows a linear positive correlation; Then, the stability of the positive correlation is judged based on historical data, where the historical data uses historical measurement data, and the measurement data includes measured water level or measured flow rate; Secondly, the mutability of the measurement data in the set time period Z is calculated randomly, and it is judged whether the measurement data in the time period Z is reliable according to the mutation rate; When the mutability is greater than the set threshold K, it is determined that the measurement data does not meet the set stability index, the measurement data in the time period Z belongs to unreliable data, and it is marked as unreliable measurement data, and the execution ends; When the mutability is not greater than the set threshold K, it is determined that the measurement data reaches the set stability index, and the measurement data in the time period Z is marked as reliable measurement data; Among them, the stability index means that the measurement data A and B are collected successively in the time period Z, The absolute value of the growth rate is abs(A - B) / A. The absolute value of the growth rate abs(A - B) / A is compared with the set growth rate. When the absolute value of the growth rate ≤ the set growth rate, it is determined that the stability index is met; When the absolute value of the growth rate > the set growth rate, it is considered that there is a mutation in the measurement data A and B, and it is determined that the stability index cannot be reached; When there are T + 1 pieces of measurement data in the time period Z, the total number of growth rates is T, and the number of mutations is C. The mutation rate is calculated as Pc = C / T. The threshold of the predefined normal range is Y. When Pc < Y, it is considered that the data in the time period Z is reliable; otherwise, there is a mutation; Secondly, based on the reliable measurement data, it is stored as historical data and a loss model is established; The loss model is a neural network model. The loss model is trained using an AI algorithm. The calibration results of the upstream and downstream measurement stations for each operation and maintenance are used as the training set. When calibrating each time, the measured values of flow rate, water level, and flow velocity are measured at the upstream and downstream measurement stations. The measured acquisition parameters of the upstream measurement station are used as reliable parameters to adjust the neural network algorithm and calculate the loss value to ensure that the measured values of flow rate, water level, and flow velocity at the downstream measurement station satisfy the upper and lower fluctuation intervals of the corresponding set logical values S2_L of flow rate, water level, and flow velocity, so as to establish a reliable model; The acquisition parameters also include time, distance between upstream and downstream measurement stations, temperature, channel type, and / or weather elements.

[0007] Furthermore, in step S2, the reliability of equipment performance is judged by using the flow balance relationship through the loss model; First, input the acquisition parameters; the acquisition parameters are used as the input values of the loss model; Then, calculate the loss value through the loss model, and infer the flow situation at the downstream measurement station; then make a judgment in combination with the measured result value at the downstream; When calculating the loss value, the downstream logical flow rate, water level, and flow velocity are simulated by inputting the collected parameters; the upper and lower fluctuation ranges are compared with the downstream measured parameters; when the downstream measured parameters are within the upper and lower fluctuation ranges, the equipment is determined to be reliable.

[0008] Furthermore, in S3, cross-validation is performed using other upstream and downstream measurement points; First, arbitrarily select at least three monitoring stations upstream and downstream; the at least three monitoring stations include, in sequence according to the flow direction, upstream station A, midstream station B, and downstream station C. Secondly, when the measured flow rate changes from station A to station B, then the flow rate at station B is not equal to the flow rate at station A; the flow rate at station C is not equal to the flow rate at station A, and the flow rate at station B is synchronized with the flow rate at station C. If the flow rate at station C is out of sync with the flow rate at station B, the equipment is deemed unreliable; where the logical flow rate at station A is Aq, the logical flow rate at station B is Bq, and the logical flow rate at station C is Cq. The deviation of measured flow at interval stations is calculated, that is, the difference between the measured flow at station A and the measured flow at station C, and is defined as the upstream and downstream measurement deviation S3_AC. Next, the range of logical flow variation is preset; taking the logical flow Aq of station A as the benchmark, the fluctuation range of the logical flow Bq of station B is calculated. If the measured flow variation of station B does not meet the fluctuation range of station B, it is determined that there is a flow imbalance, otherwise it is balanced; taking the logical flow Bq of station B as the benchmark, the flow fluctuation range of the logical flow Cq of station C is calculated. If the measured flow variation of station C does not meet the fluctuation range of station C, it is determined that there is a flow imbalance, otherwise it is balanced. Then, using the logical flow Aq of station A as a benchmark, the fluctuation range of the logical flow Cq of station C is calculated. If the measured flow change of station C does not meet the fluctuation range of station C, it is determined that there is a flow imbalance; otherwise, it is balanced.

[0009] Furthermore, in step S2, if the distance between the upstream and downstream stations is greater than the set distance, weather factors are considered, and the weather parameters of the upstream and downstream stations are input into the loss model; Weather parameters include air humidity, temperature, and / or wind speed.

[0010] Furthermore, after step S3, the following steps are performed; Step S4: Make a comprehensive judgment on steps S2 and S3 (take the weight ratio according to the actual situation) and calculate the comprehensive index R; Step S5: Judge the comprehensive index R from step S4. If it meets the set requirements, the equipment is considered reliable; otherwise, the equipment is considered unreliable.

[0011] Furthermore, in step S4, the comprehensive index R is obtained based on the logical value S2_L from step S2 and the measured deviation S3_AC between upstream and downstream measurements from step S3. The comprehensive index R = S2_L × α + S3_AC × β; Where α is the weight of S2_L; β is the weight of S3_AC; α and β are set based on experience.

[0012] Furthermore, in step S5, a threshold for the comprehensive index R is set; If the overall index R is within the set threshold range, the equipment is considered reliable.

[0013] Furthermore, a system for demonstrating the performance of flow measurement equipment using the flow balance method includes establishing a loss model using the aforementioned method, which is installed in the waterway. The loss model includes monitoring stations set up upstream and downstream in the waterway; The station is equipped with a flow meter, a water level gauge, and / or a thermometer.

[0014] Existing technologies typically involve on-site calibration, where a result is obtained manually or using other equipment and then compared with the result from a flow measurement device. This method requires calibration every time, incurring significant financial and material costs. However, the flow balance verification method of this invention avoids third-party verification. It only requires data analysis and comparison to verify the reliability of the equipment, offering advantages such as simplicity, convenience, wide applicability, and high timeliness. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating step one of the present invention.

[0016] Figure 2 This is a schematic diagram of step two of the present invention.

[0017] Figure 3 This is a flowchart illustrating step three of the present invention.

[0018] Figure 4 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0019] Example 1, as Figure 1 Step 1: Use historical data to build a loss model to determine the reliability of the equipment's own performance; First, take advantage of the characteristics of steady water flow, which is especially evident in artificial canals, where water level and / or flow rate are positively correlated, and water level will not suddenly change; the changes in water level or flow rate are linearly positively correlated. Then, the stability of the positive correlation is judged based on historical data, which includes historical measurement data such as measured water level or measured flow rate; secondly, the variability of the measurement data in a set time period Z is randomly calculated, and the reliability of the measurement data in time period Z is judged based on the variability rate. If the mutability exceeds the set threshold K, the measurement data is deemed not to meet the set stability index, and the measurement data for time period Z is considered unreliable and marked as unreliable measurement data, thus ending the execution process; if the mutability is not greater than the set threshold K, the measurement data is deemed to meet the set stability index, and the measurement data for time period Z is marked as reliable measurement data. Examples of the following definitions: The stability index is defined as follows: in the collected measurement data, the absolute value of the growth rate between two adjacent water level data points at two time points is ≤ 50% or the absolute value of the growth rate between adjacent flow data points is ≤ 50%; the data collection frequency for time period Z is set to one data point per minute. Here, we define two measurement data points, A and B, collected sequentially; then the absolute value of the growth rate is abs(AB) / A. Definition of data mutation: If the growth rate of two adjacent measurement data A and B is greater than 50%, then measurement data A and B are considered to have a mutation and are deemed to have failed to meet the stability index. Defining the reliability of data for time period Z: For example, if time period Z is 1 day, and 1 day equals 1440 minutes, then there are 1440 data points. Calculating the growth rate between two adjacent data points yields 1439 growth rates. Assuming the number of data mutations is 200, the mutation rate is calculated as: 200 / 1439 = 13.9%. If the normal range is predefined as within 5%, i.e., the threshold is Y=5%, then the data for that 1 day is considered unreliable data.

[0020] Calculation formula: In time period Z, the total growth rate is T, the number of mutations is C, and the mutation rate is Pc = C / T. When Pc < 5%, the data in time period Z is normal; otherwise, it is a mutation.

[0021] Secondly, based on reliable measurement data storage, historical data is stored and a loss model is established; The loss model is a neural network (NN) model. The loss model is trained using an AI algorithm, with the calibration results from upstream and downstream monitoring stations during each maintenance cycle serving as the training set. During each calibration, the upstream and downstream monitoring stations measure the actual values ​​of flow rate, water level, and flow velocity. The measured parameters from the upstream station are used as reliable parameters to adjust the NN algorithm and calculate the loss value. This ensures that the measured values ​​of flow rate, water level, and flow velocity at the downstream station meet the upper and lower fluctuation range requirements of the corresponding set logical value S2_L. In this case, the model is considered reliable. The upper and lower fluctuation range of the logical value S2_L can be ±20%.

[0022] Example 2, as Figure 2 Step two: Using the flow balance relationship, the reliability of the equipment performance is determined through a loss model. First, input the collected parameters; use these parameters as input values ​​for the loss model. The collected parameters include physical elements and data such as time, upstream flow rate, upstream flow velocity, upstream water level, distance between upstream and downstream stations, temperature, channel type, and weather factors. Then, the loss value is calculated using the loss model to estimate the downstream flow situation; and a judgment is made based on the results of downstream measurements. When calculating the loss value, the downstream logical flow rate, water level, and flow velocity are simulated by inputting the collected parameters. For example, if the logical flow rate L = 8 m³ / s, the judgment interval is defined as ±20%, and the downstream flow rate should be in [6.4, 9.6]. This interval is then compared with the downstream measured flow rate. Assuming the downstream measured flow rate Qc = 7, then the sign requirement is met, and the equipment is reliable.

[0023] Example 3, as Figure 3 Step 3: Cross-validate using other upstream and downstream measurement points; Arbitrarily select three monitoring stations upstream and downstream: station A, station B, and station C, with station B in the middle. Utilizing the principle of flow balance, if there is new water injection or discharge between station A and station B, then the flow rate at station B will definitely be greater than or less than the flow rate at station A, and the flow rate at station C will also definitely be greater than or less than the flow rate at station A. Station C and station B are synchronized. If station C and station B are out of sync, then there is definitely a problem, and the unreliability of the equipment performance can be determined.

[0024] The comparison here is based on the upstream station and is a range comparison. The normal variation range set by the system can be -10%. If the measured variation exceeds this normal variation range, then there may be a flow imbalance. Then, the comparison between the two downstream stations needs to be performed according to the above method to determine the problem.

[0025] For example, if the upstream station A has a logical flow rate of Aq 10 m³ / s, then the downstream station B has a logical flow rate of Bq in the interval [9, 10]. Assuming Bq = 9, then the downstream station C of station B should have a flow rate Cq in the interval [8.1, 9]. After obtaining the theoretical interval, we can then compare it with the measured values.

[0026] Secondly, cross-judgment is performed on any two sets of upstream and downstream stations (e.g., stations A and C; stations B and C; stations A and B) to determine whether the equipment is reliable.

[0027] The deviation of measured flow at interval stations is calculated, that is, the difference between the measured flow at station A and the measured flow at station C, and is defined as the upstream and downstream measurement deviation S3_AC.

[0028] Example 4, as Figure 1-4 The following description, in conjunction with Examples 1-3, serves as a general overview. Step S1: Verify the measurement results of the device itself using the steps in Example 1. If it fails, there is no need to proceed with the following steps; the device is directly determined to be unreliable. If the data mutation rate is acceptable, continue execution. Step S2: Analyze the equipment using the steps in Example 2; calculate the logical flow loss value based on the loss model using AI algorithms; then determine the reliability of the equipment by comparing it with the downstream measured values. If the downstream measured value is within the logical value range, the device is deemed reliable and execution continues; if the downstream measured value is not within the logical value range, the device is deemed unreliable and execution ends. If the upstream and downstream distances are far apart, local weather issues need to be considered. The weather conditions of the two locations need to be added to the loss model, and then a judgment is made based on the interval of the calculation results and the measured values. Step S3: Using the steps in Example 3, extract data from upstream and downstream stations, and perform cross-validation to determine the reliability of the equipment using the comparison method and threshold method. Step S4: Make a comprehensive judgment on steps S2 and S3 (take the weight ratio according to the actual situation) and calculate the comprehensive index R; Step S5: Judge the comprehensive index R from step S4. If it meets the requirements, the equipment is considered reliable; otherwise, the equipment is considered unreliable.

[0029] In step S4, the comprehensive index R is obtained based on the logical value S2_L from step S2 and the measured deviation S3_AC between upstream and downstream measurements from step S3. The overall index R = S2_L ×α + S3_AC × β; Where α is the weight of S2_L; β is the weight of S3_AC; α and β are set based on experience.

[0030] In step S5, the threshold of the comprehensive index R is set; If the overall index R is within the set threshold range, the equipment is considered reliable.

[0031] The present invention has been described in detail for the purpose of making the disclosure clearer, and the prior art will not be listed in detail.

[0032] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. It is obvious to those skilled in the art that multiple technical solutions of the present invention can be combined. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All technical contents not described in detail in the present invention are well-known technologies.

Claims

1. A method for demonstrating the performance of a flow measurement device using the flow balance method, characterized in that: It includes the following steps; Step S1: Establish a loss model using historical data to demonstrate the measurement results of the device itself. If the data mutation rate is qualified, continue; Otherwise, determine that the device is unreliable and end the execution; Step S2: Analyze the device. First, based on the loss model, calculate the logical value S2_L of the flow loss value through the AI algorithm. Then, compare the logical value S2_L with the downstream measured value to determine whether the device is reliable; If the downstream measured value is within the logical value interval corresponding to the logical value S2_L, the device is judged to be reliable and continue; if the downstream measured value is not within the logical value interval, the device is unreliable and end the execution; Step S3: Extract data from the upstream and downstream measurement stations, and perform cross-validation to judge the reliability of the device according to the comparison method and threshold method; when it is determined that the device is reliable, continue; otherwise, end the execution; The method is applied in an artificial main canal; In step S1, first, the change process of water level or flow rate shows a linear positive correlation; Then, judge the stability of the positive correlation relationship based on historical data, where the historical data uses historical measurement data, and the measurement data includes the measured water level or measured flow rate; Secondly, randomly calculate the mutability of the measurement data in the set time period Z, and judge whether the measurement data in the time period Z is reliable according to the mutation rate; When the mutability is greater than the set threshold K, it is considered that the measurement data does not meet the set stability index, and the measurement data in the time period Z belongs to unreliable and is marked as unreliable measurement data, and end the execution; When the mutability is not greater than the set threshold K, it is considered that the measurement data reaches the set stability index, and the measurement data in the time period Z is marked as reliable measurement data; Among them, the stability index refers to collecting measurement data A and B successively in the time period Z, The absolute value of the growth rate is abs(A - B) / A. Compare the absolute value of the growth rate abs(A - B) / A with the set growth rate. When the absolute value of the growth rate ≤ the set growth rate, it is judged to meet the stability index; When the absolute value of the growth rate > the set growth rate, it is considered that there is a mutation in the measurement data A and B, and it is considered that the stability index is not reached; When there are T + 1 measurement data in the time period Z, the total number of growth rates is T, and the number of mutations is C. Calculate the mutation rate Pc = C / T. The threshold of the predefined normal range is Y. When Pc < Y, it is considered that the data in the time period Z is reliable; otherwise, there is a mutation; Secondly, store the reliable measurement data as historical data and establish a loss model; The loss model is a neural network model. The loss model is trained using the AI algorithm, and the calibration results of the upstream and downstream measurement stations for each operation and maintenance are used as the training set. When calibrating each time, the measured values of flow rate, water level, and flow velocity are measured at the upstream and downstream measurement stations. The measured acquisition parameters of the upstream measurement station are used as reliable parameters, and the neural network algorithm is adjusted to calculate the loss value to ensure that the measured values of flow rate, water level, and flow velocity at the downstream measurement station satisfy the upper and lower fluctuation intervals of the corresponding flow rate, water level, and flow velocity setting logical value S2_L, so as to establish a reliable model; The collected parameters also include time, distance between upstream and downstream stations, temperature, channel type and / or weather factors; in S3, cross-validation is performed using other upstream and downstream stations; First, arbitrarily select at least three monitoring stations upstream and downstream; the at least three monitoring stations include, in sequence according to the flow direction, upstream station A, midstream station B, and downstream station C. Secondly, when the measured flow rate changes from station A to station B, then the flow rate at station B is not equal to the flow rate at station A; the flow rate at station C is not equal to the flow rate at station A, and the flow rate at station B is synchronized with the flow rate at station C. If the flow rate at station C is out of sync with the flow rate at station B, the equipment is deemed unreliable; where the logical flow rate at station A is Aq, the logical flow rate at station B is Bq, and the logical flow rate at station C is Cq. The deviation of measured flow at interval stations is calculated, that is, the difference between the measured flow at station A and the measured flow at station C, and is defined as the upstream and downstream measurement deviation S3_AC. Next, the range of logical flow variation is preset; taking the logical flow Aq of station A as the benchmark, the fluctuation range of the logical flow Bq of station B is calculated. If the measured flow variation of station B does not meet the fluctuation range of station B, it is determined that there is a flow imbalance, otherwise it is balanced; taking the logical flow Bq of station B as the benchmark, the flow fluctuation range of the logical flow Cq of station C is calculated. If the measured flow variation of station C does not meet the fluctuation range of station C, it is determined that there is a flow imbalance, otherwise it is balanced. Then, using the logical flow Aq of station A as a benchmark, the fluctuation range of the logical flow Cq of station C is calculated. If the measured flow change of station C does not meet the fluctuation range of station C, it is determined that there is a flow imbalance; otherwise, it is balanced.

2. The method for demonstrating the performance of a flow measurement device using the flow balance method according to claim 1, characterized in that: In step S2, the reliability of the equipment performance is determined by using the flow balance relationship and a loss model. First, input the collected parameters; use these parameters as input values ​​for the loss model. Then, the loss value is calculated using the loss model to estimate the flow situation at the downstream monitoring station; and a judgment is made based on the results of downstream measurements. When calculating the loss value, the downstream logical flow rate, water level, and flow velocity are simulated by inputting the collected parameters; the upper and lower fluctuation ranges are compared with the downstream measured parameters; when the downstream measured parameters are within the upper and lower fluctuation ranges, the equipment is determined to be reliable.

3. The method for demonstrating the performance of a flow measurement device using the flow balance method according to claim 2, characterized in that: In step S2, if the distance between the upstream and downstream stations is greater than the set distance, then weather factors are considered and the weather parameters of the upstream and downstream stations are input into the loss model. Weather parameters include air humidity, temperature, and / or wind speed.

4. The method for demonstrating the performance of a flow measurement device using the flow balance method according to claim 3, characterized in that: After step S3, perform the following steps; Step S4: Make a comprehensive judgment on steps S2 and S3, and calculate the comprehensive index R; Step S5: Judge the comprehensive index R from step S4. If it meets the set requirements, the equipment is considered reliable; otherwise, the equipment is considered unreliable.

5. The method for demonstrating the performance of a flow measurement device using the flow balance method according to claim 4, characterized in that: In step S4, the comprehensive index R is obtained based on the logical value S2_L from step S2 and the measured deviation S3_AC between upstream and downstream measurements from step S3. The comprehensive index R = S2_L × α + S3_AC × β; Where α is the weight of S2_L; β is the weight of S3_AC; α and β are set based on experience.

6. The method for demonstrating the performance of a flow measurement device using the flow balance method according to claim 5, characterized in that: In step S5, the threshold of the comprehensive index R is set; If the overall index R is within the set threshold range, the equipment is considered reliable.

7. A system for demonstrating the performance of flow measurement equipment using the flow balance method, characterized in that: The method described in any one of claims 1-6, which is set in a waterway, is used to build a loss model; The loss model includes monitoring stations set up upstream and downstream in the waterway; The station is equipped with a flow meter, a water level gauge, and / or a thermometer.

Images