A heating system temperature and flow sensor fault diagnosis method based on historical data

Abstract

The present application relates to a kind of heat supply system temperature, flow sensor fault diagnosis method based on historical data, the method comprises: first by temperature sensor and flow sensor collection heat station historical operation data, calculate the average heat exchange efficiency and average heat exchange coefficient of current time then the data of zero flow for real-time transmission is corrected;Finally, temperature, flow sensor is analyzed and diagnosed for fault.The above-mentioned temperature, flow sensor fault diagnosis method, not only can the heat station data of real-time monitoring be judged and analyzed, also can error and correction in time when temperature, flow sensor appears fault, greatly reduce the redundancy quantity of temperature, flow sensor, and need not real-time acquisition exception data, only when it is identified that no fault occurs, it needs to be acquired, can greatly improve the effective utilization of system resources.

Description

[Technical Field]

[0001] This invention relates to the field of fault diagnosis, and in particular to a method for diagnosing faults in temperature and flow sensors of a heating system based on historical data. [Background Technology]

[0002] The earliest method used for sensor fault diagnosis was hardware redundancy, which is the simplest and most effective, but it is costly, difficult to promote, and can only detect faults, not provide timely emergency feedback. Currently, sensor fault detection methods are generally divided into direct and indirect methods. Direct methods rely on various input and output parameters of the heating system as fault symptoms, which are fed into a classifier. The classifier classifies the fault according to a pre-defined classification strategy. However, this method is ineffective for dynamic systems and some minor faults, and has significant limitations. Indirect methods require first establishing system models under normal and known fault conditions, and then using these models to predict the system's performance. Methods for building these models include regression, neural networks, and fuzzy logic. The key to indirect methods lies in the reliability of simulation data and the accuracy of the model. If the model error is large, it will inevitably lead to misdiagnosis or missed diagnosis. For example, while neural networks have learning capabilities and good diagnostic robustness, their large learning load and slow convergence speed make them unsuitable for online learning and diagnosis.

[0003] Heating systems typically have complete temperature, pressure, and flow sensors in both primary and secondary networks at heating stations to monitor their operational status. Some smart heating systems even have advanced information systems such as DCS, SCADA, and GIS support. With the support of these advanced information technologies, heating systems can analyze and process collected historical data. Therefore, this patent proposes a method for fault diagnosis based on heating data monitoring data by establishing a physical model and processing large amounts of historical data. This method does not require establishing a precise mathematical model of the system, making development and application relatively convenient. It has low requirements for historical data and system information and possesses some "intelligent" characteristics. Furthermore, the knowledge-based quantitative model is simple, intuitive, visual, and convenient to represent; the reasoning is transparent; the diagnostic reasoning speed is fast; the required data storage space is relatively small; and the fault diagnosis results are highly interpretable. [Summary of the Invention]

[0004] To address the aforementioned problems in the prior art, this invention proposes a fault diagnosis method for temperature and flow sensors in a heating system based on historical data, the method comprising:

[0005] Step 1: Collect historical operating data of the heating station using temperature and flow sensors, and calculate the average heat exchange efficiency at the current moment. and average heat transfer coefficient

[0006] Temperature and flow sensors are used to record the following data from the previous month (every 10 minutes): primary network supply water temperature T1, primary network return water temperature T2, secondary network supply water temperature T4, secondary network return water temperature T3, primary network flow rate M12, and secondary network flow rate M34, and these data are stored in the historical database. The heat transfer coefficient K and heat transfer efficiency η are calculated and stored in the historical database. The specific calculation method is as follows:

[0007] Q1 = c × M12 × (T1 - T2)

[0008] Q2 = c × M34 × (T4 - T3)

[0009] η×Q1=Q2

[0010]

[0011] In the formula, T1 is the supply water temperature of the primary heating network, T2 is the return water temperature of the primary heating network, T3 is the return water temperature of the secondary heating network, T4 is the supply water temperature of the secondary heating network, M12 is the flow rate of the primary heating network, and M34 is the flow rate of the secondary heating network. When the heat exchanger flows in the same direction as the heat exchanger, Δt′=T1-T3, Δt″=T2-T4; when the heat exchanger flows in the opposite direction to the heat exchanger, Δt′=T1-T4, Δt″=T2-T3. In engineering practice, the average temperature difference can also be used. Instead of logarithmic mean temperature difference

[0012] Sort the calculated heat transfer coefficient K and heat transfer efficiency η for all time points, and use the 15% and 85% points of all data as the heat transfer efficiency threshold η for this heating station. min η max and heat transfer coefficient threshold K min and K max ;

[0013] Calculate the average heat transfer efficiency at the current moment. With average heat transfer coefficient The specific calculation method is as follows:

[0014]

[0015]

[0016] In the formula, η1-η 12 The heat transfer efficiency, K1-K, is calculated for the corresponding time points in the first two hours. 12 The heat transfer coefficients are calculated for the corresponding time points within the first two hours. The calculated heat transfer efficiency for each corresponding time point must be within the threshold range (η). min η max Otherwise, calculate the average heat transfer efficiency. The heat transfer efficiency calculated at that specific moment is not accepted. The calculated heat transfer coefficient for the corresponding moment must be within the threshold range (K). min K max Otherwise, calculate the average heat transfer coefficient. The heat transfer coefficient calculated at that moment is not adopted.

[0017] Step 2: Correct data with zero real-time transmission traffic.

[0018] When both the primary network flow rate M12 and the secondary network flow rate M34 of the heating station are 0 kg / s, the heating station can be considered to be in a closed state at that moment, and no correction is required. If one side of the primary network flow rate M12 or the secondary network flow rate M34 is 0 kg / s and the other side is not 0 kg / s, the flow sensor can be considered to be malfunctioning at that moment, and M12 or M34 should be corrected accordingly. The specific method is as follows:

[0019] If M12 = 0 and M34 = 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0020] If M12 = 0 and M34 × (T4 - T3) > 0, then correct M12. And read M12 and M34 into the database;

[0021] If M12 = 0 and M34 × (T4 - T3) < 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0022] If M34 = 0 and M12 × (T1 - T2) > 0, then correct M34. And read M12 and M34 into the database;

[0023] If M34 = 0 and M12 × (T1 - T2) > 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0024] Step 3: Fault Analysis and Diagnosis of Temperature Sensor and Flow Sensor

[0025] First, let's consider the heat exchange efficiency at the current moment. Make a judgment; if the heat exchange efficiency η is within... If the temperature and flow sensor readings are within the specified range, the current readings are correct and no correction is needed; the data will then be read into the database. If the heat exchange efficiency η is not within the specified range... If the reading is within the specified range, then the current meter reading may be problematic.

[0026] Diagnostic tests were performed on the following parameters: primary network supply temperature T1, primary network return temperature T2, secondary network supply temperature T4, secondary network return temperature T3, primary network flow rate M12, and secondary network flow rate M34. The specific method is as follows:

[0027] Q1 = c × M12 × (T1 - T2)

[0028] Q2 = c × M34 × (T4 - T3)

[0029] η×Q1=Q2

[0030]

[0031] To diagnose the water supply temperature T1 of the network, use the above formula to calculate the assumed T1′ and the assumed heat exchange. The system compares the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it determines that the water supply temperature T1 of the primary network has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the water supply temperature T1 of the primary network is accurate and no correction is required.

[0032] To diagnose the return water temperature T2, use the above formula to calculate the assumed T2′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the primary network return water temperature T2 has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the primary network return water temperature T2 is accurate and no correction is required.

[0033] To diagnose the secondary water supply temperature T4, use the above formula to calculate the assumed T4′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4′ - T3). Compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary water supply temperature T4 has malfunctioned and needs to be corrected. If the absolute error between the two is not less than 1, the secondary water supply temperature T4 is accurate and no correction is needed.

[0034] To diagnose the secondary network return water temperature T3, use the above formula to calculate the assumed T3′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4 - T3′), and compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network return water temperature T3 has a fault and should be corrected; if the absolute error between the two is not less than 1, the secondary network return water temperature T3 is accurate and no correction is required.

[0035] To diagnose the flow rate M12 of a single network, use the above formula to calculate the assumed M12′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the flow rate M12 of the network has failed and is corrected. If the absolute error between the two is not less than 1, the flow rate M12 of the network is accurate and no correction is required.

[0036] To diagnose the secondary network flow rate M34, use the above formula to calculate the assumed M34′ and the assumed heat exchange rate Q′2 = 4200 × M34′ × (T4 - T3). Compare the assumed heat exchange rate with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network flow rate M34 has a fault and should be corrected. If the absolute error between the two is not less than 1, the secondary network flow rate M34 is accurate and no correction is required.

[0037] The beneficial effects of this invention include: by establishing a physical model and processing large amounts of historical data, fault diagnosis can be performed on heating data monitoring data, improving the accuracy of monitoring. Furthermore, this diagnostic method breaks through the traditional method of monitoring sensor faults using hardware. This method is widely applicable in practice, does not require establishing a precise mathematical model of the system, is relatively easy to develop and apply, has low requirements for historical data and system information, and possesses some "intelligent" features. In addition, the knowledge-based quantitative model is simple, intuitive, visual, and convenient to represent; the reasoning is transparent; the diagnostic reasoning speed is fast; the required data storage space is relatively small; and the fault diagnosis results are highly interpretable. [Attached Image Description]

[0038] Figure 1 This is a schematic diagram of the primary and secondary network systems of the heating station according to the present invention;

[0039] Figure 2 This is a flowchart illustrating the calculation process for correcting real-time transmitted data with zero traffic, as per the present invention.

[0040] Figure 3 This is a flowchart illustrating the calculation process for fault diagnosis of temperature sensors and flow sensors according to the present invention.

Detailed Implementation Methods

[0041] The fault diagnosis method for temperature and flow sensors provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0042] Step 1: Collect historical operating data of the heating station using temperature and flow sensors, and calculate the average heat exchange efficiency at the current moment. and average heat transfer coefficient

[0043] Temperature and flow sensors are used to record the following data from the previous month (every 10 minutes): primary network supply water temperature T1, primary network return water temperature T2, secondary network supply water temperature T4, secondary network return water temperature T3, primary network flow rate M12, and secondary network flow rate M34, and these data are stored in the historical database. The heat transfer coefficient K and heat transfer efficiency η are calculated and stored in the historical database. The specific calculation method is as follows:

[0044] Q1 = c × M12 × (T1 - T2)

[0045] Q2 = c × M34 × (T4 - T3)

[0046] η×Q1=Q2

[0047]

[0048] In the formula, T1 is the supply water temperature of the primary heating network, T2 is the return water temperature of the primary heating network, T3 is the return water temperature of the secondary heating network, T4 is the supply water temperature of the secondary heating network, M12 is the flow rate of the primary heating network, and M34 is the flow rate of the secondary heating network. When the heat exchanger flows in the same direction as the heat exchanger, Δt′=T1-T3, Δt″=T2-T4; when the heat exchanger flows in the opposite direction to the heat exchanger, Δt′=T1-T4, Δt″=T2-T3. In engineering practice, the average temperature difference can also be used. Instead of logarithmic mean temperature difference

[0049] Sort the calculated heat transfer coefficient K and heat transfer efficiency η for all time points, and use the 15% and 85% points of all data as the heat transfer efficiency threshold η for this heating station. min η max and heat transfer coefficient threshold K min and K max ;

[0050] Calculate the average heat transfer efficiency at the current moment. With average heat transfer coefficient The specific calculation method is as follows:

[0051]

[0052]

[0053] In the formula, η1-η 12 The heat transfer efficiency, K1-K, is calculated for the corresponding time points in the first two hours. 12 The heat transfer coefficients are calculated for the corresponding time points within the first two hours. The calculated heat transfer efficiency for each corresponding time point must be within the threshold range (η). min η max Otherwise, calculate the average heat transfer efficiency. The heat transfer efficiency calculated at that specific moment is not accepted. The calculated heat transfer coefficient for the corresponding moment must be within the threshold range (K). min K max Otherwise, calculate the average heat transfer coefficient. The heat transfer coefficient calculated at that moment is not adopted.

[0054] Step 2: Correct data with zero real-time transmission traffic.

[0055] When both the primary network flow rate M12 and the secondary network flow rate M34 of the heating station are 0 kg / s, the heating station can be considered to be in a closed state at that moment, and no correction is required. If one side of the primary network flow rate M12 or the secondary network flow rate M34 is 0 kg / s and the other side is not 0 kg / s, the flow sensor can be considered to be malfunctioning at that moment, and M12 or M34 should be corrected accordingly. The specific method is as follows:

[0056] If M12 = 0 and M34 = 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0057] If M12 = 0 and M34 × (T4 - T3) > 0, then correct M12. And read M12 and M34 into the database;

[0058] If M12 = 0 and M34 × (T4 - T3) < 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0059] If M34 = 0 and M12 × (T1 - T2) > 0, then correct M34. And read M12 and M34 into the database;

[0060] If M34 = 0 and M12 × (T1 - T2) > 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database;

[0061] Step 3: Fault Analysis and Diagnosis of Temperature Sensor and Flow Sensor

[0062] First, let's consider the heat exchange efficiency at the current moment. Make a judgment; if the heat exchange efficiency η is within... If the temperature and flow sensor readings are within the specified range, the current readings are correct and no correction is needed; the data will then be read into the database. If the heat exchange efficiency η is not within the specified range... If the reading is within the specified range, then the current meter reading may be problematic.

[0063] Diagnostic tests were performed on the following parameters: primary network supply temperature T1, primary network return temperature T2, secondary network supply temperature T4, secondary network return temperature T3, primary network flow rate M12, and secondary network flow rate M34. The specific method is as follows:

[0064] Q1 = c × M12 × (T1 - T2)

[0065] Q2 = c × M34 × (T4 - T3)

[0066] η×Q1=Q2

[0067]

[0068] To diagnose the water supply temperature T1 of the network, use the above formula to calculate the assumed T1′ and the assumed heat exchange. The system compares the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it determines that the water supply temperature T1 of the primary network has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the water supply temperature T1 of the primary network is accurate and no correction is required.

[0069] To diagnose the return water temperature T2, use the above formula to calculate the assumed T2′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the primary network return water temperature T2 has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the primary network return water temperature T2 is accurate and no correction is required.

[0070] To diagnose the secondary water supply temperature T4, use the above formula to calculate the assumed T4′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4′ - T3). Compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary water supply temperature T4 has malfunctioned and needs to be corrected. If the absolute error between the two is not less than 1, the secondary water supply temperature T4 is accurate and no correction is needed.

[0071] To diagnose the secondary network return water temperature T3, use the above formula to calculate the assumed T3′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4 - T3′), and compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network return water temperature T3 has a fault and should be corrected; if the absolute error between the two is not less than 1, the secondary network return water temperature T3 is accurate and no correction is required.

[0072] To diagnose the flow rate M12 of a single network, use the above formula to calculate the assumed M12′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the flow rate M12 of the network has failed and is corrected. If the absolute error between the two is not less than 1, the flow rate M12 of the network is accurate and no correction is required.

[0073] To diagnose the secondary network flow rate M34, use the above formula to calculate the assumed M34′ and the assumed heat exchange rate Q′2 = 4200 × M34′ × (T4 - T3). Compare the assumed heat exchange rate with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network flow rate M34 has a fault and should be corrected. If the absolute error between the two is not less than 1, the secondary network flow rate M34 is accurate and no correction is required.

[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for fault diagnosis of temperature and flow sensors in a heating system based on historical data, characterized in that... Includes the following steps: Step 1: Collect historical operating data of the heating station using temperature and flow sensors, and calculate the average heat exchange efficiency at the current moment. and average heat transfer coefficient Temperature and flow sensors are used to record the following data from the previous month (every 10 minutes): primary network supply water temperature T1, primary network return water temperature T2, secondary network supply water temperature T4, secondary network return water temperature T3, primary network flow rate M12, and secondary network flow rate M34, and these data are stored in the historical database. The heat transfer coefficient K and heat transfer efficiency η are calculated and stored in the historical database. The specific calculation method is as follows: Q1 = c × M12 × (T1 - T2) Q2 = c × M34 × (T4 - T3) η×Q1=Q2 In the formula, T1 is the supply water temperature of the primary heating network, T2 is the return water temperature of the primary heating network, T3 is the return water temperature of the secondary heating network, T4 is the supply water temperature of the secondary heating network, M12 is the flow rate of the primary heating network, and M34 is the flow rate of the secondary heating network. When the heat exchanger flows in the same direction as the heat exchanger, Δt′=T1-T3, Δt″=T2-T4; when the heat exchanger flows in the opposite direction to the heat exchanger, Δt′=T1-T4, Δt″=T2-T3. In engineering practice, the average temperature difference can also be used. Instead of logarithmic mean temperature difference Sort the calculated heat transfer coefficient K and heat transfer efficiency η for all time points, and use the 15% and 85% points of all data as the heat transfer efficiency threshold η for this heating station. min η max and heat transfer coefficient threshold K min and K max ; Calculate the average heat transfer efficiency at the current moment. With average heat transfer coefficient The specific calculation method is as follows: In the formula, η1-η 12 The heat transfer efficiency, K1-K, is calculated for the corresponding time points in the first two hours. 12 The heat transfer coefficients are calculated for the corresponding time points within the first two hours. The calculated heat transfer efficiency for each corresponding time point must be within the threshold range (η). min η max Otherwise, calculate the average heat transfer efficiency. The heat transfer efficiency calculated at that specific moment is not accepted. The calculated heat transfer coefficient for the corresponding moment must be within the threshold range (K). min K max Otherwise, calculate the average heat transfer coefficient. The heat transfer coefficient calculated at that moment is not adopted. Step 2: Correct data with zero real-time transmission traffic. When both the primary network flow rate M12 and the secondary network flow rate M34 of the heating station are 0 kg / s, the heating station can be considered to be in a closed state at that moment, and no correction is required. If one side of the primary network flow rate M12 or the secondary network flow rate M34 is 0 kg / s and the other side is not 0 kg / s, the flow sensor can be considered to be malfunctioning at that moment, and M12 or M34 should be corrected accordingly. The specific method is as follows: If M12 = 0 and M34 = 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database; If M12 = 0 and M34 × (T4 - T3) > 0, then correct. And read M12 and M34 into the database; If M12 = 0 and M34 × (T4 - T3) < 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database; If M34 = 0 and M12 × (T1 - T2) > 0, then correct M34. And read M12 and M34 into the database; If M34 = 0 and M12 × (T1 - T2) > 0, then read M12 = 0 kg / s and M34 = 0 kg / s into the database; Step 3: Fault Analysis and Diagnosis of Temperature Sensor and Flow Sensor First, let's consider the heat exchange efficiency at the current moment. Make a judgment; if the heat exchange efficiency η is within... If the temperature and flow sensor readings are within the specified range, the current readings are correct and no correction is needed; the data will then be read into the database. If the heat exchange efficiency η is not within the specified range... If the reading is within the specified range, then the current meter reading may be problematic. Diagnostic tests were performed on the following parameters: primary network supply temperature T1, primary network return temperature T2, secondary network supply temperature T4, secondary network return temperature T3, primary network flow rate M12, and secondary network flow rate M34. The specific method is as follows: Q1 = c × M12 × (T1 - T2) Q2 = c × M34 × (T4 - T3) η×Q1=Q2 To diagnose the water supply temperature T1 of the network, use the above formula to calculate the assumed T1′ and the assumed heat exchange. The system compares the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it determines that the water supply temperature T1 of the primary network has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the water supply temperature T1 of the primary network is accurate and no correction is required. To diagnose the return water temperature T2, use the above formula to calculate the assumed T2′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the primary network return water temperature T2 has malfunctioned and is corrected. If the absolute error between the two is not less than 1, the primary network return water temperature T2 is accurate and no correction is required. To diagnose the secondary water supply temperature T4, use the above formula to calculate the assumed T4′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4′ - T3). Compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary water supply temperature T4 has malfunctioned and needs to be corrected. If the absolute error between the two is not less than 1, the secondary water supply temperature T4 is accurate and no correction is needed. To diagnose the secondary network return water temperature T3, use the above formula to calculate the assumed T3′ and the assumed heat exchange Q′2 = 4200 × M34 × (T4 - T3′), and compare the assumed heat exchange with the heat exchange at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network return water temperature T3 has a fault and should be corrected; if the absolute error between the two is not less than 1, the secondary network return water temperature T3 is accurate and no correction is required. To diagnose the flow rate M12 of a single network, use the above formula to calculate the assumed M12′ and the assumed heat exchange. The assumed heat exchange rate is compared with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the flow rate M12 of the network has failed and is corrected. If the absolute error between the two is not less than 1, the flow rate M12 of the network is accurate and no correction is required. To diagnose the secondary network flow rate M34, use the above formula to calculate the assumed M34′ and the assumed heat exchange rate Q′2 = 4200 × M34′ × (T4 - T3). Compare the assumed heat exchange rate with the heat exchange rate at the current moment. If the absolute error between the two is less than 1, it is determined that the secondary network flow rate M34 has a fault and should be corrected. If the absolute error between the two is not less than 1, the secondary network flow rate M34 is accurate and no correction is required.

Images