Boiler pressure vessel inspection and detection system and method

Abstract

The invention belongs to the technical field of pressure vessel safety monitoring, and particularly relates to a boiler pressure vessel inspection and detection system and method based on multi-sensor fusion, and the system comprises a pressure sensor array which is arranged at a plurality of detection points on the inner wall of a boiler vessel and is used for collecting pressure data in real time; the temperature compensation module is connected with the pressure sensor array and used for correcting a pressure measurement value according to the environment temperature; the wireless transmission unit is used for sending the corrected pressure data to a remote monitoring terminal; and the early warning analysis module is used for establishing a pressure change model based on the historical data and triggering early warning when the real-time data deviates from a model threshold value. And the detection precision is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of pressure vessel safety monitoring, and specifically provides a boiler pressure vessel inspection and detection system and method based on multi-sensor fusion, which is suitable for real-time safety monitoring of pressure-bearing equipment such as power station boilers and industrial boilers. Background Art

[0002] Boiler pressure vessels are special equipment with explosion hazards. Boiler pressure vessels, the full name for both boilers and pressure vessels, are primarily used in chemical and petrochemical processes for heat transfer, mass transfer, reaction, and other processes, as well as for the storage and transportation of pressurized gases or liquefied gases. They are also widely used in other industrial and civilian fields. Boiler pressure vessels are enclosed devices that contain gas or liquid and carry a certain pressure. Pressure vessels require testing after manufacturing.

[0003] A boiler pressure vessel inspection and testing device disclosed in Chinese patent CN219956836U can rotate the rotating disk 180 degrees to rotate the installed boiler to the position directly below the pressure detection assembly for inspection. At the same time, the inspected boiler can be disassembled, thereby improving the efficiency of inspection. However, while solving the problem, the boiler pressure vessel inspection and testing device has the following defects: it is not convenient to seal the container, which affects the inspection effect, and the inspection angle is single, the result reliability is low, and the device is not stable, which affects its use.

[0004] Through the study of traditional boiler pressure detection, the following defects exist:

[0005] 1. Using a single-point pressure gauge for monitoring cannot reflect the overall stress state of the container. The disclosed mechanical pressure gauge has a measurement error of ±2.5%;

[0006] 2. The manual inspection cycle is long, and it is difficult to detect sudden pressure fluctuations in time;

[0007] 3. Existing early warning systems only compare static thresholds and do not consider the temperature-pressure coupling effect;

[0008] 4. There is a lack of linkage analysis between pressure data and container wall thickness loss. Summary of the Invention

[0009] The purpose of the present invention is to provide a boiler pressure vessel inspection and detection system and method based on multi-sensor fusion to address the problems existing in the prior art. The system adopts an "axial + circumferential" three-dimensional sensor layout strategy, which greatly improves the detection accuracy.

[0010] The technical solution of the present invention is:

[0011] A boiler container pressure inspection and detection system, comprising

[0012] Pressure sensor arrays are arranged at multiple detection points on the inner wall of the boiler vessel to collect pressure data in real time;

[0013] a temperature compensation module, connected to the pressure sensor array, for correcting the pressure measurement value according to the ambient temperature;

[0014] The wireless transmission unit sends the corrected pressure data to the remote monitoring terminal;

[0015] The early warning analysis module establishes a pressure change model based on historical data and triggers an early warning when real-time data deviates from the model threshold.

[0016] Specifically, the pressure sensor array uses at least three groups of MEMS pressure sensors, which are circumferentially arranged in the middle of the boiler container at intervals of 120 degrees; each group includes 3-5 sensor nodes distributed along the axial direction.

[0017] Specifically, the temperature compensation module includes:

[0018] PT100 temperature sensor mounted on the outer wall of the boiler container;

[0019] Compensation algorithm based on material thermal expansion coefficient: ΔP = α·(T-T0)·P0, where α is the compensation coefficient, T is the real-time temperature, and T0 is the reference temperature.

[0020] Specifically, the warning analysis module establishes a pressure-time change prediction model through an LSTM neural network; a first-level warning is activated when three consecutive measured values exceed the predicted value within ±5%; and a second-level emergency warning is activated when the instantaneous pressure change rate exceeds 10kPa / s.

[0021] Specifically, it also includes:

[0022] Laser thickness measuring unit, synchronously detecting the wall thickness of boiler container;

[0023] The safety factor calculation module calculates the real-time safety factor according to the formula S = (P·D) / (2σ·t), where P is the measured pressure, D is the container diameter, σ is the material yield strength, and t is the measured wall thickness.

[0024] A method for testing and inspecting the pressure of a boiler container comprises the following steps:

[0025] S1. Collect multi-dimensional pressure data while the boiler is operating;

[0026] S2. Perform temperature drift compensation and vibration noise filtering on the raw pressure data;

[0027] S3. Compare the processed data with a preset safety pressure curve;

[0028] S4. When pressure anomalies are detected, a diagnostic report including the abnormality location and risk level is automatically generated.

[0029] Specifically, in step S2 of the above boiler container pressure inspection method, the following steps are adopted:

[0030] Noise filtering based on wavelet transform;

[0031] Adaptive Kalman filter algorithm based on boiler historical operation data.

[0032] Specifically, the temperature drift compensation in step S2 of the above boiler container pressure inspection and detection method includes the following steps:

[0033] Get the pressure sensor reading P_raw and real-time wall temperature T;

[0034] Select linear or nonlinear compensation mode according to temperature range;

[0035] Perform thermal stress coupling compensation calculations.

[0036] Specifically, the implementation of the above-mentioned adaptive Kalman filter algorithm based on historical boiler operation data includes the following steps:

[0037] Establish a noise parameter feature library through historical operation data;

[0038] The Q / R matrix is updated online using the variational Bayesian method;

[0039] Automatic enhanced tracking mode is activated when a sudden change in boiler load is detected.

[0040] Specifically, the triggering conditions for the automatic enhanced tracking mode described above are:

[0041] The residual χ of three consecutive sampling periods 2 Value > 6.0;

[0042] Boiler load change rate>8% / min.

[0043] The beneficial effects of the present invention are as follows: the boiler vessel pressure inspection and detection system provided by the present invention adopts an "axial + circumferential" three-dimensional sensor layout strategy, uses high-temperature resistant silicon piezoresistive sensors, and has an operating temperature range of -40°C to 300°C; the detection accuracy is improved to ±0.1% FS; a hybrid compensation algorithm that integrates temperature compensation and vibration filtering algorithms maintains measurement stability under temperature mutation conditions; and uses a pressure-wall thickness correlation method based on machine learning to predict the remaining service life. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a schematic diagram of the sensor array arrangement;

[0045] Figure 2 It is a flow chart of the compensation algorithm based on the thermal expansion coefficient of the material.

[0046] 1MEMS pressure sensor node location, 2boiler shell. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and specific implementation methods.

[0048] Implementation Method 1

[0049] This embodiment provides a boiler container pressure inspection and detection system, including

[0050] Pressure sensor arrays are arranged at multiple detection points on the inner wall of the boiler vessel to collect pressure data in real time;

[0051] a temperature compensation module, connected to the pressure sensor array, for correcting the pressure measurement value according to the ambient temperature;

[0052] The wireless transmission unit sends the corrected pressure data to the remote monitoring terminal. The wireless transmission is based on the LoRaWAN protocol and the transmission distance is ≥500m.

[0053] The early warning analysis module establishes a pressure change model based on historical data and triggers an early warning when real-time data deviates from the model threshold.

[0054] The pressure sensor array in this embodiment uses three groups of MEMS pressure sensors. The pressure sensors are Honeywell 26PCB series with a range of 0-10 MPa. They use Inconel718 alloy brackets with a temperature resistance of 650°C. The sensors are arranged circumferentially in the middle of the boiler container at 120° intervals. Each group contains three sensor nodes distributed along the axial direction, such as Figure 1 shown.

[0055] The temperature compensation module includes:

[0056] PT100 temperature sensor mounted on the outer wall of the boiler container;

[0057] Compensation algorithm based on material thermal expansion coefficient: ΔP = α·(T-T0)·P0, where α is the compensation coefficient, T is the real-time temperature, and T0 is the reference temperature.

[0058] In this embodiment, the warning analysis module establishes a pressure-time change prediction model through an LSTM neural network; when three consecutive measured values exceed the predicted value within ±5%, a level one yellow warning is activated; when the instantaneous pressure change rate exceeds 10 kPa / s, a level two orange emergency warning is activated.

[0059] To further improve detection accuracy, the enhanced design in this embodiment also includes:

[0060] Laser thickness measuring unit, synchronously detecting the wall thickness of boiler container;

[0061] The safety factor calculation module calculates the real-time safety factor according to the formula S = (P·D) / (2σ·t), where P is the measured pressure, D is the container diameter, σ is the material yield strength, and t is the measured wall thickness. A red alert is activated when the early warning analysis module detects that the safety factor S is less than 1.5.

[0062] Example 2

[0063] This embodiment provides a method for testing boiler container pressure based on the system provided in Example 1, including the following steps:

[0064] S1. Collect multi-dimensional pressure data while the boiler is operating;

[0065] S2. Perform temperature drift compensation and vibration noise filtering on the raw pressure data;

[0066] S3. Compare the processed data with a preset safety pressure curve;

[0067] S4. When pressure anomalies are detected, a diagnostic report including the abnormality location and risk level is automatically generated.

[0068] The difference in thermal expansion between the sensor housing and the boiler material (e.g. the α difference between the 304 stainless steel housing and the carbon steel boiler is 3.2×10 -6 The additional strain error caused by thermal stress under high temperature conditions (measured up to 2-5% of the rated value) is 1.0447 / °C. Therefore, the temperature drift compensation in step S2 of the boiler container pressure inspection method provided in this embodiment includes the following steps:

[0069] Get the pressure sensor reading P_raw and real-time wall temperature T;

[0070] Select linear or nonlinear compensation mode according to temperature range;

[0071] Perform thermal stress coupling compensation calculations.

[0072] This compensation method achieves precise compensation through three levels of correction: basic linear compensation: linear correction based on the standard thermal expansion coefficient α; nonlinear segment compensation: for the nonlinear expansion characteristics of materials above 300°C; stress coupling compensation: eliminates the cross-influence of thermal stress and working pressure.

[0073] The compensation formula is as follows:

[0074] ΔP_{comp}=P_{raw}×[α_1(T-T_0)+β(T-T_0)^2]-γ×\frac{EΔT}{D}×t

[0075] in:

[0076] α_1: linear expansion coefficient (1.2×10 -5 / ℃ for carbon steel)

[0077] β: nonlinear correction coefficient (5.8×10 -9 / ℃ 2 , determined by material testing)

[0078] γ: Poisson's ratio correlation factor (0.3-0.35)

[0079] E: elastic modulus (200GPa)

[0080] D: Boiler diameter (m)

[0081] t: wall thickness (m)

[0082] A[original pressure P_raw]-->B{temperature T≤300℃}

[0083] B--Yes-->C[Linear compensation: P_comp=P_raw×α1ΔT]

[0084] B--No-->D[Nonlinear compensation: P_comp=P_raw×α1ΔT+βΔT 2 ]

[0085] C&D-->E[Stress compensation: P_final=P_comp-γEΔT / D×t]

[0086] E-->F[Output pressure after compensation]

[0087] In this embodiment, the implementation of the adaptive Kalman filter algorithm based on the boiler historical operation data includes the following steps:

[0088] Establish a noise parameter feature library through historical operation data;

[0089] The Q / R matrix is updated online using the variational Bayesian method;

[0090] Automatic enhanced tracking mode is activated when a sudden change in boiler load is detected.

[0091] The triggering conditions for the automatic enhanced tracking mode described above are:

[0092] The residual χ of three consecutive sampling periods 2 Value > 6.0;

[0093] Boiler load change rate>8% / min.

[0094] In the step S2 of the above-mentioned boiler container pressure inspection and detection method of this embodiment, noise filtering processing based on wavelet transform and adaptive Kalman filtering algorithm based on historical boiler operation data are adopted.

[0095] The above adaptive Kalman filter algorithm is dynamically optimized through the following improvements: historical data driven Q / R adaptive **: using the past 30 days of operating data to build a noise statistical feature library; variational Bayesian learning: online update of the covariance matrix of process noise Q and observation noise R; fault mode recognition: when the residual χ 2 Automatically switches to strong tracking mode when the test exceeds the threshold.

[0096] The state equation is as follows:

[0097] x_k=A·x_{k-1}+B·u_k+w_k

[0098] in:

[0099] x_k=[pressure P; pressure change rate dP / dt] is the state vector

[0100] A=[1Δt;01] (Δt is the sampling interval)

[0101] w_k~N(0,Q) process noise

[0102] Adaptive Q matrix update rule:

[0103] Q_k=α·Q_{base}+(1-α)·(K_k·z_k·z_k^T·K_k^T)

[0104] in:

[0105] α=0.95 is the forgetting factor

[0106] Q_{base} is obtained from historical data statistics

[0107] K_k is the Kalman gain

[0108] z_k is the observation residual.

[0109] This embodiment performs offline training on the above algorithm on a 350MW unit boiler as follows:

[0110] 1. Collect boiler pressure data in the 20%-100% load range;

[0111] 2. Use the EM algorithm to solve the optimal Q and R parameters under each working condition;

[0112] 3. Establish a load-noise parameter mapping table (see Table 1)

[0113] Table 1

[0114] Load rate Q(11) Q(22) R 20％ 0.01 0.001 0.05 50％ 0.03 0.002 0.03 100％ 0.05 0.005 0.01

[0115] 4. Real-time detection of boiler load change rate δ;

[0116] 5. When δ>5% / min, start the strong tracking algorithm:

[0117] K_k=P_k·H^T·[H·P_k·H^T+R_k]^{-1}×(1+γ·||z_k||)

[0118] Where γ = 0.1 is the tracking gain coefficient;

[0119] 6. Perform variational Bayes parameter updates every 10 minutes.

[0120] The above tests show that: - The sensitivity of pressure fluctuation peak detection is improved by 42%; the filter convergence time is shortened from 15s of the traditional method to 3.2s; and the mean square error under load mutation conditions is reduced to 0.008MPa.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or some technical features can be replaced by equivalents without departing from the spirit of the technical solution of the present invention. They should all be included in the scope of the technical solution for protection of the present invention.

Claims

1. A boiler container pressure inspection and detection system, characterized in that: include Pressure sensor arrays are arranged at multiple detection points on the inner wall of the boiler vessel to collect pressure data in real time; a temperature compensation module, connected to the pressure sensor array, for correcting the pressure measurement value according to the ambient temperature; The wireless transmission unit sends the corrected pressure data to the remote monitoring terminal; The early warning analysis module establishes a pressure change model based on historical data and triggers an early warning when real-time data deviates from the model threshold.

2. The boiler container pressure inspection and detection system according to claim 1, characterized in that: The pressure sensor array uses at least three groups of MEMS pressure sensors, which are evenly spaced and circumferentially arranged in the middle of the boiler container; each group includes 3-5 sensor nodes distributed along the axial direction.

3. The boiler container pressure inspection and detection system according to claim 1, characterized in that: The temperature compensation module includes: PT100 temperature sensor mounted on the outer wall of the boiler container; Compensation algorithm based on material thermal expansion coefficient: ΔP = α·(T-T0)·P0, where α is the compensation coefficient, T is the real-time temperature, and T0 is the reference temperature.

4. The boiler container pressure inspection and detection system according to claim 1, characterized in that: The warning analysis module establishes a pressure-time change prediction model through an LSTM neural network; a first-level warning is activated when three consecutive measured values exceed the predicted value within ±5%; and a second-level emergency warning is activated when the instantaneous pressure change rate exceeds 10kPa / s.

5. The boiler container pressure inspection and detection system according to claim 1, characterized in that: Also includes: Laser thickness measuring unit, synchronously detecting the wall thickness of boiler container; The safety factor calculation module calculates the real-time safety factor according to the formula S = (P·D) / (2σ·t), where P is the measured pressure, D is the container diameter, σ is the material yield strength, and t is the measured wall thickness.

6. A boiler container pressure inspection and detection method, characterized in that: The steps include: S1. Collect multi-dimensional pressure data while the boiler is operating; S2. Perform temperature drift compensation and vibration noise filtering on the raw pressure data; S3. Compare the processed data with a preset safety pressure curve; S4. When pressure anomalies are detected, a diagnostic report including the abnormal location and risk level is automatically generated.

7. The boiler container pressure inspection and detection method according to claim 6, characterized in that: In the step S2, the following is adopted: Noise filtering based on wavelet transform; Adaptive Kalman filter algorithm based on boiler historical operation data.

8. The boiler container pressure inspection and detection method according to claim 6, characterized in that: The temperature drift compensation in step S2 includes the following steps: Get the pressure sensor reading P_raw and real-time wall temperature T; Select linear or nonlinear compensation mode according to temperature range; Perform thermal stress coupling compensation calculations.

9. The boiler container pressure inspection and testing method according to claim 7, characterized in that: The implementation of the adaptive Kalman filter algorithm based on historical boiler operation data includes the following steps: Establish a noise parameter feature library through historical operation data; The Q / R matrix is updated online using the variational Bayesian method; Automatic enhanced tracking mode is activated when a sudden change in boiler load is detected.

10. The boiler container pressure inspection and testing method according to claim 9, characterized in that: The triggering conditions of the automatic enhanced tracking mode are: The residual χ2 value for three consecutive sampling periods is greater than 6.0; Boiler load change rate>8% / min.

Images