High-pressure valve flow capacity anomaly detection and strength failure early warning system

Abstract

The application relates to the field of anomaly detection, and discloses a high-pressure valve flow capacity anomaly detection and strength failure early warning system, which is used for accurately evaluating the thermodynamic and kinetic states of a valve. The system comprises the following steps: obtaining original parameters by synchronously measuring the fluid pressure, temperature, mass flow and valve body vibration acceleration of the inlet and outlet of the valve; calculating an entropy production health index for representing the abnormal degree of energy dissipation in the valve; simultaneously calculating a chaos index for representing the instability degree of the structure kinetic state; nonlinearly fusing the two indexes to generate a composite health index with higher sensitivity and an evolution track of the composite health index; constructing a random dynamic system evolution model based on the track; and projecting to obtain the failure probability of the future state of the valve. The application avoids relying on historical fault data and machine learning models, realizes deep measurement, accurate testing and intuitive indication of the health state of the high-pressure valve through chain processing driven by physical mechanism.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of anomaly detection, and particularly relates to a high-pressure valve flow capacity anomaly detection and strength failure early warning system. BACKGROUND

[0002] Supercritical once-through reheat technology is one of the most efficient and advanced clean coal power generation technologies in the modern coal-fired power generation field. In this unit, high-pressure valves, especially main steam valves, regulating valves, and high-pressure heater three-way valves, are the core key equipment for controlling super-high temperature and super-high pressure steam flow, ensuring the safety of unit start-stop and flexible peak shaving.

[0003] During the long-term operation of high-pressure valves, they are easily affected by various complex factors such as fluid scouring, temperature changes, pressure fluctuations, and mechanical vibrations, which can cause flow capacity anomalies and strength failures. Flow capacity anomalies can lead to unstable fluid flow, affecting the precise control of the production process, while strength failures can cause valve leakage, rupture, and other serious accidents, resulting in personnel injuries, environmental pollution, and huge economic losses.

[0004] Currently, there has been some research on health monitoring and fault warning technology for high-pressure valves. Traditional methods mainly rely on single physical parameter monitoring, such as assessing the running state of the valve by monitoring only individual parameters such as pressure, temperature, or vibration.

[0005] These methods have obvious limitations:

[0006] Single parameters are difficult to fully and accurately reflect the complex thermodynamic and kinetic state changes of the valve. The failure of a high-pressure valve is often the result of multiple factor coupling, and it is difficult to accurately determine the type and severity of the failure based on the changes of individual parameters.

[0007] Changes in pressure and temperature can be caused by various reasons, including changes in fluid properties, internal structural damage of the valve, and external working condition fluctuations, so it is impossible to distinguish the specific reasons by only these two parameters.

[0008] Traditional methods lack the ability to model the dynamic evolution of the valve's health state and predict the failure probability. Existing monitoring technologies can only display real-time running parameters, and cannot predict the future health state of the valve, making it difficult to achieve early warning and providing insufficient time and basis for maintenance decisions.

[0009] Therefore, we propose a high-pressure valve flow capacity anomaly detection and strength failure early warning system to solve the above problems. SUMMARY

[0010] The application provides a high-pressure valve flow capacity anomaly detection and strength failure early warning system for accurately evaluating the thermodynamic and kinetic state of the valve.

[0011] The first aspect of the application provides a high-pressure valve flow capacity anomaly detection and strength failure early warning system, comprising: an acquisition module for synchronously measuring the fluid pressure and temperature, fluid mass flow at the inlet and outlet of the high-pressure valve, and vibration acceleration of the valve body; a processing module for obtaining an entropy production health index based on the thermodynamic law according to the acquired fluid pressure, temperature and fluid mass flow; a setting module for obtaining a chaos index for characterizing the instability degree of the valve structure kinetic state based on the nonlinear time series analysis technology according to the vibration acceleration; a fusion module for fusing the entropy production health index and the chaos index, generating a composite health index, and taking the change sequence of the composite health index over time as a system state evolution characteristic trajectory; and a distribution module for constructing a random dynamic system evolution model according to the system state evolution characteristic trajectory, projecting to obtain the failure probability of the future state of the valve, so as to realize the early warning of the flow capacity anomaly and strength failure risk.

[0012] Optionally, in the first implementation manner of the first aspect of the application, the method comprises: collecting a first pressure measurement value and a first temperature measurement value of the fluid at the inlet of the high-pressure valve, collecting a second pressure measurement value and a second temperature measurement value of the fluid at the outlet of the high-pressure valve, collecting a real-time mass flow measurement value of the fluid flowing through the high-pressure valve, and collecting a vibration acceleration analog signal of the surface of the valve body of the high-pressure valve; performing analog-digital conversion on the vibration acceleration analog signal to generate a vibration acceleration digital signal sequence sampled at equal time intervals; for each data point in the first pressure measurement value, the first temperature measurement value, the second pressure measurement value, the second temperature measurement value, the real-time mass flow measurement value and the vibration acceleration digital signal sequence, a time stamp is correspondingly generated to form a multi-parameter synchronous measurement data set.

[0013] Optionally, in the second implementation form of the first aspect of the present application, the method comprises: obtaining a first pressure measurement value, a second pressure measurement value, a first temperature measurement value, a second temperature measurement value and a real-time mass flow measurement value in the multi-parameter synchronous measurement data set; determining a first specific entropy value of fluid at the valve inlet based on the first pressure measurement value and the first temperature measurement value by querying a pre-stored high-precision fluid property parameter database; determining a second specific entropy value of fluid at the valve outlet based on the second pressure measurement value and the second temperature measurement value by querying the high-precision fluid property parameter database; calculating an absolute entropy generation rate of the valve under the current operating state based on the real-time mass flow measurement value, the first specific entropy value and the second specific entropy value; obtaining a preset minimum entropy generation rate of the valve under a health benchmark state, and generating an entropy generation health index by comparing the absolute entropy generation rate with the minimum entropy generation rate.

[0014] Optionally, in the third implementation form of the first aspect of the present application, the actual entropy generation is :

[0015] ;

[0016] wherein, is the inlet specific entropy, is the outlet specific entropy, is the mass flow;

[0017] ;

[0018] is the entropy generation health index.

[0019] Optionally, in the fourth implementation form of the first aspect of the present application, the method comprises: reconstructing a phase space of a vibration acceleration digital signal sequence in the multi-parameter synchronous measurement data set, determining an optimal delay time by calculating a self-correlation function or a mutual information function, determining an optimal embedding dimension by calculating a false nearest neighbor ratio, and generating a multi-dimensional phase space trajectory; tracking a long-time evolution behavior of neighboring points on the phase space trajectory in the reconstructed multi-dimensional phase space to obtain a maximum Lyapunov exponent; processing the obtained maximum Lyapunov exponent, comparing the maximum Lyapunov exponent with a reference Lyapunov exponent benchmark value representing a system background noise level, and generating a chaos index for quantitatively representing a degree of instability of a current dynamic state relative to a stable benchmark.

[0020] Optionally, in the fifth implementation form of the first aspect of the present application, the method further comprises: dynamically calculating a real-time weighting factor according to statistical dispersion characteristics of the real-time time series of the entropy generation health indicator and the real-time time series of the chaos indicator; inputting the real-time weighting factor, an instantaneous value of the entropy generation health indicator and an instantaneous value of the chaos indicator into a preset nonlinear function to generate an instantaneous composite health indicator value; and collecting continuous instantaneous composite health indicator values in chronological order to form a composite health indicator time series trajectory representing a valve health state degradation path.

[0021] Optionally, in the sixth implementation form of the first aspect of the present application, the method further comprises: identifying a drift coefficient and a diffusion coefficient of a stochastic differential equation based on the composite health indicator time series trajectory by calculating conditional statistical moments thereof at different time delays; constructing a Fokker-Planck equation describing evolution of a composite health indicator probability density function over time based on the identified drift coefficient and diffusion coefficient; setting a composite health indicator threshold value corresponding to a valve failure critical state, and obtaining a probability density distribution of the composite health indicator first reaching the failure threshold value from a current value by solving the Fokker-Planck equation, to obtain a time-varying failure probability and an average remaining useful life estimate within a specified future time period; mapping the time-varying failure probability to a first analog control signal for controlling a deflection angle of a probability gauge pointer, and mapping the average remaining useful life estimate to a second analog control signal for controlling a display length of a remaining life progress bar.

[0022] Optionally, in the seventh implementation form of the first aspect of the present application, the method further comprises a composite module: constructing a two-dimensional health state phase plane with the entropy generation health indicator and the chaos indicator as coordinate axes, defining a health stable region on the two-dimensional health state phase plane based on an index data cluster distribution under an initial health state of the valve; dynamically adjusting a boundary of the health stable region based on the time-varying failure probability, wherein a width of the boundary region is proportional to the time-varying failure probability, and mapping an instantaneous value of the composite health indicator to a visual warning attribute of a corresponding coordinate point on the two-dimensional health state phase plane.

[0023] Optionally, in the eighth implementation form of the first aspect of the present application, when the coordinate point continuously deviates from the center of the health stable region and the time-varying failure probability corresponding to the coordinate point exceeds the first preset threshold, a first-level early warning signal is generated, when the coordinate point reaches or exceeds the dynamically adjusted rear boundary of the health stable region and the time-varying failure probability corresponding to the coordinate point exceeds the second preset threshold, a second-level early warning signal is generated, and the first-level early warning signal and the second-level early warning signal are converted into an industrial communication protocol signal output; after the second-level early warning signal is generated, the motion trajectory direction of the coordinate point on the two-dimensional health state phase plane is analyzed, if the trajectory mainly expands along the entropy production health index axis direction, an auxiliary judgment signal indicating flow passage abnormality is generated, if the trajectory mainly expands along the chaos index axis direction, an auxiliary judgment signal indicating structure strength abnormality is generated, and if the trajectory expands along both axes, an auxiliary judgment signal indicating a compound failure mode is generated, and the auxiliary judgment signal and the early warning signal are synchronously output.

[0024] The mechanism of the present application is as follows: efficient conversion of measurement results into intuitive and decision-making information is achieved, thereby providing an intelligent measurement indication scheme for high-pressure valves with the capabilities of early warning and failure physical mechanism tracing;

[0025] Beneficial effects: by dynamically calculating the weighting factor, the contributions of thermodynamics and dynamics are balanced, so that the compound health index has higher sensitivity, the index can comprehensively consider the factors of valve internal energy dissipation and structure dynamics state, more comprehensively and accurately reflects the health state of the valve, and provides a more reliable basis for subsequent failure probability prediction and early warning;

[0026] The random evolution process of the valve health state can be considered, the future state can be scientifically predicted, early warning is realized, compared with the traditional method of judging based on the current state, potential risks can be found in advance, more time is obtained for maintenance decision-making, and the probability of accident occurrence is effectively reduced;

[0027] A two-dimensional health state phase plane is constructed, the compound health index is mapped into a coordinate point, and a health stable region is defined, by analyzing the motion trajectory direction of the coordinate point, the fault type can be accurately judged, failure tracing is realized, intuitive and clear fault information is provided for maintenance personnel, which is helpful for quickly locating the fault position and reason, formulating a targeted maintenance strategy, improving the maintenance efficiency, and reducing the maintenance cost. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 It is an embodiment schematic diagram of the high-pressure valve flow capacity abnormality detection and strength failure early warning system in the embodiment of the present application;

[0029] Figure 2 It is another embodiment schematic diagram of the high-pressure valve flow capacity abnormality detection and strength failure early warning system in the embodiment of the present application;

[0030] Figure 3 Figure 1 is a schematic diagram of an embodiment of the high-pressure valve flow capacity abnormality detection and strength failure early warning device in the present application. DETAILED DESCRIPTION

[0031] The high-pressure valve flow capacity abnormality detection and strength failure early warning system in the embodiments of the present application is used to accurately evaluate the thermodynamic and kinetic states of the valve. The terms "first", "second", "third", "fourth" and the like (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0032] For the sake of understanding, the specific flow of the embodiments of the present application is described below. Please refer to Figure 1 An embodiment of the high-pressure valve flow capacity abnormality detection and strength failure early warning system in the embodiments of the present application includes:

[0033] 101, an acquisition module, configured to acquire original parameters for calculating the thermodynamic and kinetic states of the valve by synchronously measuring the fluid pressure and temperature, fluid mass flow rate at the inlet and outlet of the high-pressure valve, and vibration acceleration of the valve body;

[0034] It can be understood that the execution subject of the present application can be a high-pressure valve flow capacity abnormality detection and strength failure early warning device, and can also be a terminal or a server, and the specific execution subject is not limited herein. The embodiments of the present application take the server as the execution subject for example.

[0035] It should be noted that in the high-pressure valve flow capacity abnormality detection of the supercritical double-reheat unit, the synchronous measurement of the fluid parameters at the inlet and outlet of the valve and the vibration data of the valve body is the basis for subsequent analysis:

[0036] Measurement point arrangement and parameter selection: Taking the super-high pressure main steam valve of a supercritical double-reheat unit (main steam pressure 28.3 MPa, temperature 620℃) as an example, the following are arranged at the valve inlet pipeline (1.5 times the pipe diameter away from the front end of the valve body) and the outlet pipeline (2 times the pipe diameter away from the rear end of the valve body):

[0037] High-temperature pressure sensor (range 0-40MPa, accuracy ±0.1%), measures inlet pressure (28.5MPa) and outlet pressure (28.3MPa).

[0038] Type K thermocouple (temperature range 0-800℃, accuracy ±0.5℃) measures inlet fluid temperature (605℃) and outlet fluid temperature (602℃).

[0039] Vortex flow meter (installed in the inlet straight pipe section, ensuring a straight pipe diameter of 10 times that of the pipe before and 5 times that of the pipe after) measures steam mass flow rate (98.5 kg / s).

[0040] Meanwhile, a triaxial vibration acceleration sensor (frequency response 0.5-10kHz, range ±500g) is installed at the top flange of the valve body to collect vibration acceleration signals of the valve body in three orthogonal directions.

[0041] Synchronous data acquisition and processing: All sensors are synchronously triggered through a high-speed data acquisition system (sampling rate ≥10kHz) to ensure time series consistency. During acquisition: pressure and temperature signals are converted into 4-20mA analog signals by transmitters, and then converted into digital signals by the acquisition card; vibration acceleration signals are directly output as digital quantities to avoid interference from analog transmission; 60 seconds of raw data are continuously acquired, including multiple sets of synchronous samples under the steady-state operating conditions of the unit (load 660MW).

[0042] Data verification and output: Instantaneous value alignment and unit conversion are performed on the collected data (pressure to Pa, temperature to K, flow rate to kg / s, vibration acceleration retained in m / s²). 2 The final output is a set of original parameter sequences with strictly synchronized timestamps, including:

[0043] Inlet pressure P in =28.5×10 6 Pa, outlet pressure P out =28.3×10 6 Pa, inlet temperature T in =878.15K, outlet temperature T out =875.15K, mass flow rate =98.5kg / s², vibration acceleration (effective values ​​in X / Y / Z directions are 2.8m / s²) 2 3.1m / s 2 2.5m / s 2 ).

[0044] 102. Processing module, used to calculate the entropy production health index to characterize the degree of abnormal energy dissipation inside the valve based on the acquired fluid pressure, temperature and fluid mass flow rate and the laws of thermodynamics.

[0045] It should be noted that the high-pressure main steam valve of a supercritical double-reheat unit is taken as an example to illustrate how to calculate the entropy production health index representing the degree of abnormal energy dissipation in the valve according to the synchronously measured fluid pressure, temperature and mass flow data.

[0046] Suppose that when the unit is in steady-state operation at a load of 660 MW, the original data of the high-pressure valve at a certain time are synchronously measured as follows: inlet fluid pressure P in = 28.5 MPa; outlet fluid pressure P out = 26.8 MPa; inlet fluid temperature T in = 605℃ (878.15 K); outlet fluid temperature T out = 598℃ (871.15 K); fluid mass flow rate = 98.5 kg / s (superheated steam).

[0047] The above data are converted into standard units (pressure into Pa and temperature into K), and the data validity is checked (instantaneous interference noise is eliminated).

[0048] The reference entropy change of an ideal reversible process is calculated. According to the second law of thermodynamics, the entropy changes before and after the valve in an adiabatic process (isentropic process) should be zero.

[0049] However, the actual process is irreversible, and the deviation of the actual entropy change from the reference entropy change needs to be calculated. In this example, the isentropic process under the design condition is taken as the reference.

[0050] According to the inlet parameters (P in , T in ) and the outlet pressure (P out ), the ideal outlet enthalpy value is obtained by consulting the steam thermodynamic property table (or IAPWS-97 model).

[0051] The actual outlet enthalpy value is obtained by consulting the measured outlet temperature .

[0052] The actual entropy production (a measure of irreversibility) is calculated: entropy production ; wherein and are obtained by consulting the thermodynamic table according to the measured inlet and outlet temperatures and pressures, respectively.

[0053] The entropy production health index is generated. The entropy production health index (EPHI) is defined as the relative deviation of the actual entropy production from the reference entropy production:

[0054] The reference value of the entropy production under the design condition (healthy state) is = 0.15 kW / K (determined by historical health data statistics).

[0055] This example calculates the actual entropy production. =0.21kW / K.

[0056] but .

[0057] An increase in this indicator indicates that energy dissipation inside the valve is intensified, which may be due to increased flow resistance (wear, scaling) or seal failure leading to increased leakage flow.

[0058] The indicator output and anomaly criteria output the EPHI value of the continuous time series and set the threshold (±20% is the normal range).

[0059] In this example, EPHI=40% exceeds the threshold, triggering a warning signal and indicating abnormal valve flow capacity (increased internal throttling effect or aggravated flow separation).

[0060] 103. Setting module, used to calculate the chaotic index to characterize the degree of instability of the valve structure dynamic state based on vibration acceleration and nonlinear time series analysis technology;

[0061] It should be noted that, taking the high-pressure main steam valve of a supercritical double reheat unit as an example, this paper explains how to calculate the chaotic index characterizing the degree of instability of the valve structure dynamic state based on the valve body vibration acceleration data using nonlinear time series analysis technology.

[0062] During steady-state operation at 660MW load, 60 seconds of vibration data were collected using a triaxial vibration accelerometer (sampling rate 10kHz) mounted on the top of the valve body. A 5-second frame of the X-direction acceleration time series (a total of 50,000 points) was selected for analysis.

[0063] The raw acceleration data is in m / s². 2 Typical values ​​range from ±15 m / s 2 Between these steps, the data undergoes detrending and filtering (using a 0.5-2kHz bandpass filter) to eliminate low-frequency drift and high-frequency noise interference.

[0064] Nonlinear feature extraction (Lyapunov exponent calculation) was performed, and the nonlinear characteristics of the vibration signal were analyzed using the phase space reconstruction method: a time delay of τ = 0.5 ms was chosen (determined based on the first zero-crossing point of the signal autocorrelation function). The embedding dimension m = 6 (calculated using the spurious nearest neighbor method). The reconstructed phase space trajectory showed that the system exhibited a sensitive dependence (initial small deviations amplified exponentially with iteration).

[0065] The maximum Lyapunov exponent (λ) was calculated to quantify the degree of chaos in the system: using the Wolf algorithm, λ = 0.85 bits / s was obtained (positive values ​​are typical of chaotic systems). For comparison, the historical baseline value λ0 = 0.12 bits / s under healthy valve conditions was obtained (derived from statistical data of similar valves operating without faults).

[0066] Chaos index generation and interpretation: The Chaotic Health Index (CHI) is defined as the relative deviation of the current Lyapunov index from the benchmark value.

[0067] ;

[0068] This indicator significantly exceeds the normal threshold (set at ±50%), indicating severe instability in the valve's structural dynamics. Possible causes include valve stem jamming, loose internal components, or nonlinear resonance induced by fluid excitation forces. A time-varying sequence of CHI is output (calculated every 5 seconds) for subsequent integration with thermodynamic indicators and failure warning.

[0069] 104. Fusion module, used to fuse entropy production health indicators and chaos indicators to generate a more sensitive composite health indicator for comprehensively indicating the health status of valves, and use the change sequence of this composite health indicator over time as the system state evolution characteristic trajectory.

[0070] It should be noted that, taking the high-pressure main steam valve of a supercritical double reheat unit as an example, this paper explains how to integrate the entropy production health index (EPHI) and the chaos index (CHI) to generate a composite health index (CHI) that comprehensively reflects the health status of the valve, and constructs its system state evolution characteristic trajectory over time.

[0071] Input indicator preparation and normalization: Assuming continuous monitoring for 5 minutes at a unit load of 660MW, one set of indicators is acquired per minute:

[0072] Entropy Production Health Index (EPHI) (characterizing thermodynamic energy dissipation): baseline health value 0.15 kW / K, actual monitored values ​​[0.18, 0.21, 0.25, 0.28, 0.30] kW / K, calculated relative deviation sequence [20%, 40%, 67%, 87%, 100%].

[0073] Chaos index (CHI) (characterizing the degree of dynamic instability): The baseline Lyapunov exponent is 0.12 bits / s, the actual calculated value is [0.35, 0.85, 1.20, 1.50, 1.80] bits / s, and the relative deviation sequence is [192%, 608%, 900%, 1150%, 1400%].

[0074] The min-max normalization is performed on both indicator sequences to eliminate the dimension effect and map the values to the range of 0-1 (0 represents health, and 1 represents severe abnormality).

[0075] The composite health indicator is generated by weighted fusion, with weights assigned according to valve characteristics: EPHI reflects internal flow efficiency (weight 0.6), and CHI reflects mechanical structural stability (weight 0.4).

[0076] The composite health indicator (CHI) calculation formula is:

[0077] ;

[0078] Taking the data at the 3rd minute as an example:

[0079] The normalized EPHI is 0.67 (corresponding to a 67% deviation), and the normalized CHI is 0.72 (corresponding to a 900% deviation).

[0080] CHI=(0.67×0.6)+(0.72×0.4)=0.402+0.288=0.69.

[0081] The CHI sequence calculation result within 5 minutes is: [0.31, 0.58, 0.69, 0.82, 0.93].

[0082] The system state evolution characteristic trajectory is constructed by aligning the CHI sequence with the timestamp (time interval 1 minute), generating a time-health state evolution curve: T1=0.31, T2=0.58, T3=0.69, T4=0.82, T5=0.93.

[0083] The trajectory shows that the CHI value continuously increases (from 0.31 to 0.93), indicating that the valve health state is accelerating deterioration, and the trend line slope gradually increases.

[0084] Abnormal threshold and early warning triggering: set two threshold values: mild abnormality threshold (CHI=0.6): trigger preliminary warning at the 2nd minute; severe abnormality threshold (CHI=0.8): trigger high-level warning at the 4th minute.

[0085] Trajectory analysis shows that the CHI value exceeds the historical baseline (health benchmark CHI<0.2) after the 3rd minute, indicating that the valve may have a rising comprehensive failure risk due to internal fouling (thermodynamic abnormality) and component loosening (dynamic instability).

[0086] The CHI time series is output as the system state evolution characteristic trajectory to the next link (Fokker-Planck model) for failure probability prediction.

[0087] At the same time, a visual trend chart is generated to help operation and maintenance personnel monitor the valve health degradation rate in real time and support predictive maintenance decisions.

[0088] 105. An allocation module is configured to construct a random dynamic system evolution model based on the Fokker-Planck equation according to the system state evolution characteristic trajectory, project the failure probability of the future state of the valve, and convert it into a control signal to drive a visual indication device to realize early warning of abnormal flow capacity and intensity failure risk.

[0089] It should be noted that the high-pressure main steam valve of a supercritical double-reheat unit is taken as an example to illustrate how to construct a random dynamic system evolution model based on the Fokker-Planck equation according to the time series of the composite health index (CHI) (system state evolution characteristic trajectory), project the failure probability of the future state of the valve, and generate a visual warning signal.

[0090] Input data: Assume that the time series of the composite health index (CHI) within 5 minutes is [0.31, 0.58, 0.69, 0.82, 0.93] (one point per minute), and the CHI value range is 0-1 (0 represents health, and 1 represents severe failure).

[0091] Random dynamics modeling: The CHI sequence is regarded as the observation value of a random process, and a probability density evolution model is constructed based on the Fokker-Planck equation. This equation describes the law of the probability distribution of the CHI value changing over time, and its form is:

[0092] ;

[0093] where is the probability density function of the CHI value, and represent the drift coefficient and the diffusion coefficient, respectively, which are obtained by fitting historical data.

[0094] Parameter estimation: The drift term (trend term) and the diffusion term (fluctuation term) are calculated according to the CHI sequence. In this example, the linear drift coefficient μ = 0.15 min −1 (representing the CHI rising rate) and the diffusion coefficient σ = 0.05 min −1 / 2 (representing the random fluctuation amplitude) are obtained by fitting.

[0095] Failure probability projection, define failure threshold: Set the failure threshold of CHI as D = 0.95 (above this value, the valve is considered to be about to fail).

[0096] Solving for the future probability distribution: The Fokker-Planck equation is used to solve for the probability distribution of CHI at future time points. Starting from the current time (minute 5), the evolution within the next 30 minutes is predicted: 10 minutes later (t=15 minutes), the mean of the probability distribution of CHI is 0.31 + 0.15 × 10 = 1.81 (but due to boundary constraints), the actual calculated probability density is concentrated in the high value region;

[0097] Calculate the probability integral of failure as the CHI value exceeding the threshold D=0.95:

[0098] ;

[0099] Specific numerical results: After 5 minutes (t=10 minutes): Failure probability =15%;

[0100] After 15 minutes (t=20 minutes): the failure probability rises to 52%;

[0101] After 30 minutes (t=35 minutes): the failure probability reaches 92% (close to inevitable failure).

[0102] Control signal generation and visual early warning, signal conversion: mapping failure probability to 4-20mA analog control signal (industrial standard):

[0103] <20%: Output 4mA (normal state, green indicator light);

[0104] 20%≤ <50%: Output 8-12mA (mild warning, yellow indicator light flashing);

[0105] 50%≤ <80%: Output 12-16mA (moderate warning, orange indicator light stays on);

[0106] ≥80%: Output 20mA (emergency warning, red indicator light stays on + buzzer alarm).

[0107] This example demonstrates the procedure based on the 30-minute prediction results. =92%, the system outputs a 20mA signal, triggering the highest level warning (red alert), prompting maintenance personnel to immediately intervene for inspection or shutdown for repair.

[0108] Visualization extension: The CHI trajectory curve and failure probability curve are displayed on the real-time monitoring interface, and the predicted remaining lifespan is marked (in this example, when the failure probability exceeds 80%, the remaining lifespan is about 15 minutes).

[0109] In this embodiment of the invention, the acquisition module simultaneously measures multiple dimensions of raw parameters, such as fluid pressure, temperature, mass flow rate, and valve body vibration acceleration, comprehensively covering key information on the valve's thermodynamic and dynamic state. Compared to traditional single-parameter detection methods, this approach more accurately reflects the valve's actual operating condition, providing a rich and accurate data foundation for subsequent analysis. The processing and setting modules respectively calculate entropy production health indicators and chaos indicators, evaluating the valve's state from two different dimensions: thermodynamic energy dissipation and structural dynamic instability. The fusion module combines these two indicators to generate a composite health indicator, comprehensively considering factors such as the valve's internal flow efficiency and mechanical structural stability. This significantly improves the accuracy and comprehensiveness of valve health status detection, effectively avoiding misjudgments and omissions that may occur with single-indicator analysis. It can acquire valve operating data in real time and quickly calculate various indicators, promptly detecting problems such as abnormal valve flow capacity and structural dynamic instability. For example, in the calculation of entropy production health indicators, by setting thresholds, an early warning signal can be triggered immediately when the indicators exceed the normal range. This allows maintenance personnel to take timely measures to prevent further escalation of the fault and reduce unit downtime and economic losses caused by valve failures. Furthermore, by constructing a stochastic dynamic system evolution model, the failure probability of valves in the future can be predicted based on the time series of composite health indicators. This proactive early warning function enables maintenance personnel to develop maintenance plans in advance, rationally schedule maintenance time, and perform preventative maintenance before serious valve failures occur, greatly improving equipment reliability and availability and reducing the impact of sudden failures on production.

[0110] Please see Figure 2 Another embodiment of the high-pressure valve flow capacity abnormality detection and strength failure early warning system in this invention includes:

[0111] 201. Acquisition module, used to acquire raw parameters for calculating the thermodynamic and dynamic states of the valve by synchronously measuring the fluid pressure and temperature, fluid mass flow rate and vibration acceleration of the valve body at the inlet and outlet of the high-pressure valve;

[0112] Specifically, the system collects the first pressure and first temperature measurements of the fluid at the inlet of the high-pressure valve; the second pressure and second temperature measurements of the fluid at the outlet of the high-pressure valve; the real-time mass flow rate of the fluid flowing through the high-pressure valve; and the simulated vibration acceleration signal of the valve body surface. The simulated vibration acceleration signal is then converted from analog to digital to generate a sequence of digital vibration acceleration signals sampled at equal time intervals. A unified, high-precision timestamp is generated for each data point in all the first pressure measurements, first temperature measurements, second pressure measurements, second temperature measurements, real-time mass flow rate measurements, and the digital vibration acceleration signal sequence, forming a time-aligned multi-parameter synchronous measurement dataset.

[0113] It should be noted that the construction process of the multi-parameter synchronous measurement dataset is illustrated by taking the Z-type high-pressure regulating valve (nominal diameter DN200, design pressure 25MPa, design temperature 540℃) on the main steam pipeline of a 600MW unit in a thermal power plant as an example.

[0114] A high-precision pressure transmitter (range 0-40MPa, accuracy 0.1%FS) and a type K armored thermocouple (range 0-800℃, accuracy ±1.5℃) are installed 2D upstream of the valve inlet (D is the pipe diameter) to collect the first inlet pressure measurement (24.85MPa) and the first inlet temperature measurement (538.7℃) in real time. A pressure transmitter and thermocouple of equal precision are installed 5D downstream of the valve outlet to collect the second outlet pressure measurement (18.36MPa) and the second temperature measurement (536.2℃). A vortex mass flow meter (range 0-300t / h, accuracy 0.5%) is installed on the valve outlet pipe to collect the real-time steam mass flow measurement (215.6t / h).

[0115] An ICP-type triaxial accelerometer (range ±500g, frequency range 0.5-10kHz) is installed on the upper surface of the valve body (vibration-sensitive area) to collect analog vibration acceleration signals. Analog-to-digital conversion is performed using a 24-bit high-resolution data acquisition card (sampling rate 20kHz) to generate a sequence of digital vibration acceleration signals sampled at equal time intervals (50μs), effectively capturing high-frequency dynamic characteristics such as valve core vibration and fluid excitation.

[0116] All sensor signals are accessed through a distributed acquisition system with a unified time base, and a unified timestamp is generated by a high-precision temperature-controlled crystal oscillator (clock deviation <1ppm). The system uses the PTP1588 precision clock protocol to synchronize the time of each acquisition node (microsecond-level synchronization accuracy), ensuring strict time alignment of inlet pressure (24.85MPa@10:05:32.123456), outlet pressure (18.36MPa@10:05:32.123456), flow rate (215.6t / h@10:05:32.123456), and vibration data points (-0.152g@10:05:32.123456), ultimately forming a multi-parameter synchronous measurement dataset.

[0117] 202. Processing module, used to calculate the entropy production health index, which characterizes the degree of abnormal energy dissipation inside the valve, based on the acquired fluid pressure, temperature and fluid mass flow rate and the laws of thermodynamics.

[0118] Specifically, the system acquires the first pressure measurement, second pressure measurement, first temperature measurement, second temperature measurement, and real-time mass flow rate measurement from a multi-parameter synchronous measurement dataset. Based on the fluid's physical properties, and using the first pressure and first temperature measurements, it determines the first specific entropy value of the fluid at the valve inlet by querying a pre-stored high-precision fluid property parameter database. Based on the fluid's physical properties, and using the second pressure and second temperature measurements, it determines the second specific entropy value of the fluid at the valve outlet by querying the high-precision fluid property parameter database. According to the second law of thermodynamics, based on the real-time mass flow rate measurement, the first specific entropy value, and the second specific entropy value, it calculates the absolute entropy production rate of the valve under its current operating state. It acquires the minimum entropy production rate of the valve under a healthy baseline state, which is pre-determined through high-fidelity simulation or initial valve health state measurement. It then compares the absolute entropy production rate with the minimum entropy production rate to generate a dimensionless entropy production health index that quantitatively characterizes the degree of deviation of the current energy dissipation from the healthy baseline.

[0119] It should be noted that the calculation process of entropy production health indicators is illustrated by taking the high-pressure regulating valve (design pressure 25MPa, design temperature 540℃) on the main steam pipeline of a 600MW unit in a thermal power plant as an example.

[0120] Based on the synchronous measurement dataset, the inlet parameters (pressure 24.85 MPa, temperature 538.7℃), outlet parameters (pressure 18.36 MPa, temperature 536.2℃), and mass flow rate (215.6 t / h) at a certain moment were obtained. According to the steam physical property parameter database (IAPWS-IF97 standard), the inlet specific entropy was found to be 6.428 kJ / (kg·K) and the outlet specific entropy was 6.512 kJ / (kg·K) through interpolation.

[0121] The absolute entropy yield is calculated according to the second law of thermodynamics. The absolute entropy yield is obtained by multiplying the mass flow rate by the entropy difference between the inlet and outlet: Absolute entropy yield = 215.6 × 1000 / 3600 kg / s × (6.512 − 6.428) kJ / (kg·K) = 5.02 kW / K. This value characterizes the actual energy dissipation level of the valve at present.

[0122] The minimum entropy production rate under this operating condition was determined to be 4.35 kW / K using a high-fidelity fluid simulation model (assuming an ideal state with no valve wear and no scaling). The dimensionless entropy production health index was generated by comparing the absolute entropy production rate with the minimum entropy production rate: Entropy production health index = 5.02 / 4.35 ≈ 1.15;

[0123] This value indicates that the current energy dissipation deviates from the healthy baseline by 15%, which may be caused by increased internal turbulence due to wear or scaling in the flow channels.

[0124] 203. Setting module, used to calculate the chaotic index to characterize the degree of instability of the valve structure dynamic state based on vibration acceleration and nonlinear time series analysis technology;

[0125] Specifically, the vibration acceleration digital signal sequence from the multi-parameter synchronous measurement dataset is acquired; the phase space of the vibration acceleration digital signal sequence is reconstructed, and the optimal delay time is determined by calculating its autocorrelation function or mutual information function, and the optimal embedding dimension is determined by calculating the proportion of spurious nearest neighbors, so as to generate a multi-dimensional phase space trajectory that is topologically equivalent to the original dynamic system; in the reconstructed multi-dimensional phase space, the long-term evolution behavior of neighboring points on the phase space trajectory is tracked, and the maximum Lyapunov exponent is calculated based on the small data quantity method. This exponent is used to quantify the sensitivity dependence of the system's dynamic trajectory on initial conditions and the degree of chaos; the calculated maximum Lyapunov exponent is dimensionless and compared with the reference Lyapunov exponent benchmark value that characterizes the system's background noise level to generate a dimensionless chaotic index used to quantitatively characterize the degree of instability of the current dynamic state relative to the stable benchmark;

[0126] It should be noted that the calculation process of the chaos index is illustrated by taking the high-pressure regulating valve (design pressure 25MPa, design temperature 540℃) on the main steam pipeline of a 600MW unit in a thermal power plant as an example.

[0127] Vibration acceleration digital signal sequences (sampling rate 20kHz, duration 10 seconds, total 200,000 data points) were extracted from a multi-parameter synchronous measurement dataset. A segment of steady-state data (2000 points between timestamps 10:05:32.123 and 10:05:32.133) was selected, with acceleration values ​​fluctuating between -2.5g and +2.8g (1g = 9.8 m / s²). 2 ).

[0128] The optimal delay time τ was determined by calculating the first zero-crossing point of the autocorrelation function of the sequence, using 15 sampling points (0.75 ms). The optimal embedding dimension m=5 was calculated using the false nearest neighbor method. According to Takens' embedding theorem, the 5-dimensional phase space trajectory was reconstructed using the delay time τ and the embedding dimension m, generating a total of 1985 phase points.

[0129] In the reconstructed phase space, an initial phase point P0 is selected (coordinates: [-0.12g, 0.35g, -0.28g, 0.15g, -0.40g]), and its nearest neighbor P1 is found (Euclidean distance d0 = 0.018g). The evolution of the two points is tracked with the number of iterations (500 iterations in total), and the relative distance d after each iteration is calculated. i .

[0130] The Rosenstein algorithm (a method using a small amount of data) is used to fit the linear region of the logarithmic distance sequence ln(dᵢ) and the number of iterations; the slope of this region is the maximum Lyapunov exponent λ. max ≈0.45 bits / sample (meaning that adjacent orbitals diverge at an exponential rate e^(0.45t)).

[0131] Obtain the reference Lyapunov index λ for the valve in a healthy state.

[0132] ref =0.12 bits / sample (calculated or simulated from initial fault-free vibration data). For λ max

[0133] Perform dimensionless processing:

[0134] ;

[0135] This value indicates that the current valve dynamics are 275% more chaotic than the healthy baseline, which may be caused by structural instability due to valve core wear, loose connections, or internal fluid excitation.

[0136] 204. Fusion module, used to fuse entropy production health indicators and chaos indicators to generate a more sensitive composite health indicator for comprehensively indicating the health status of valves, and use the change sequence of the composite health indicator over time as the system state evolution characteristic trajectory.

[0137] Specifically, the real-time time series of the dimensionless entropy production health index generated by the thermodynamic state index calculation step is obtained; the real-time time series of the dimensionless chaotic index generated by the kinetic state index calculation step is obtained; based on the statistical dispersion characteristics of the entropy production health index time series and the chaotic index time series, a real-time weighting factor for balancing thermodynamic and kinetic contributions is dynamically calculated; the real-time weighting factor, the instantaneous value of the entropy production health index, and the instantaneous value of the chaotic index are input into a preset nonlinear function to calculate and generate an instantaneous composite health index value that comprehensively reflects the overall health status of the valve; the continuous instantaneous composite health index values ​​are collected in chronological order to form a composite health index time series trajectory that can characterize the degradation path of the valve's health status.

[0138] It should be noted that the calculation process of the composite health index is illustrated using the high-pressure regulating valve on the main steam pipeline of a 600MW unit in a thermal power plant as an example. Assume that the entropy production health index obtained at the current moment is 1.15 (representing a 15% deviation of energy dissipation from the health baseline), and the chaos index is 3.75 (representing a 275% increase in the degree of dynamic chaos compared to the baseline).

[0139] Dynamic weighting factor calculation was used to analyze the statistical characteristics of the entropy production health index time series (the most recent 100 data points): its coefficient of variation (standard deviation / mean) was 0.18, indicating that the thermodynamic state changes were relatively stable.

[0140] Analyzing the statistical characteristics of the time series of the chaos index (100 data points in the same period): its coefficient of variation is 0.52, indicating that the dynamic state fluctuates significantly.

[0141] Based on the inverse relationship between the coefficients of variation of the two factors, the real-time weighting factor is dynamically calculated:

[0142] ;

[0143] ;

[0144] (Weight allocation principle: Indicators with higher volatility are more sensitive and therefore should be assigned higher weights to enhance the response to anomalies.)

[0145] The nonlinear fusion function is calculated using a weighted geometric mean (preset function form):

[0146] ;

[0147] Substitute the instantaneous value:

[0148] ;

[0149] This value comprehensively reflects the overall health status of the valve: a value greater than 1 indicates a deviation from the health benchmark, and the higher the value, the greater the risk of failure.

[0150] Time series trajectory generation involves continuously calculating composite values ​​in chronological order (one point every 10 seconds) to form a composite health index time series. The values ​​for five consecutive time points are: 1.45, 1.52, 1.58, 1.63, and 1.70. This series clearly shows the continuous degradation path of the valve's health status (a monotonically increasing trend), providing an evolutionary trajectory for subsequent predictions.

[0151] In this case, the composite value (1.58) is significantly higher than the healthy baseline (1.0), and the entropy production index (1.15) and chaos index (3.75) are both abnormal, suggesting that the valve may have a composite failure mode of both flow channel wear (increased thermodynamic energy dissipation) and structural loosening (increased kinetic chaos).

[0152] This composite index is more sensitive than a single index: if we only look at the entropy production index (1.15), the deviation is not significant, but after integrating the chaos index, the overall abnormal amplitude is amplified to 58%, triggering an early warning earlier.

[0153] 205. Allocation module, used to construct a stochastic dynamic system evolution model based on the Fokker-Planck equation according to the system state evolution characteristic trajectory, project the failure probability of the valve's future state and convert it into a control signal to drive the visual indicator device, so as to realize the early warning of abnormal flow capacity and strength failure risk.

[0154] Specifically, the process involves obtaining the time series trajectory of the composite health index generated by the composite health index generation step; based on the time series trajectory of the composite health index, the drift coefficient and diffusion coefficient of the stochastic differential equation describing the stochastic evolution process of the composite health index are identified online by calculating its conditional statistical moments under different time delays; based on the identified drift coefficient and diffusion coefficient, a Fokker-Planck equation describing the evolution of the probability density function of the composite health index over time is constructed; a threshold for the composite health index corresponding to the critical state of valve failure is set, and the probability density distribution of the composite health index reaching the failure threshold for the first time from its current value is calculated by solving the Fokker-Planck equation, thereby deriving the time-varying failure probability and the estimated average remaining useful life within a specified future time period; the time-varying failure probability is mapped to a first analog control signal for controlling the deflection angle of a probability gauge pointer, and the estimated average remaining useful life is mapped to a second analog control signal for controlling the display length of a remaining useful life progress bar.

[0155] It should be noted that the high-pressure regulating valve on the main steam pipeline of a 600MW unit in a thermal power plant is used as an example to illustrate the failure early warning implementation process based on the Fokker-Planck equation. It is assumed that a composite health index time series (100 recent data points) has been generated through previous steps, and its value has gradually increased from 1.0 (health baseline) to the current value of 1.58, showing a clear degradation trend.

[0156] Based on the time series of a composite health indicator, its conditional statistical moments (first and second moments) within the time delay interval [0, 10Δt] (Δt = 10 seconds) are calculated. By fitting using the least squares method, the parameters of the stochastic differential equation describing the stochastic evolution of this indicator are identified online.

[0157] Drift coefficient μ = 0.025 / hour (characterizing the deterministic trend of index degradation over time);

[0158] The diffusion coefficient σ = 0.18 / √h (characterizing the intensity of fluctuations caused by random disturbances);

[0159] The Fokker-Planck equation is constructed and solved based on the identified parameters (μ=0.025, σ=0.18). The Fokker-Planck equation describing the evolution of the probability density function of the composite health index is then constructed.

[0160] ;

[0161] Where z is the composite health index value. Its probability density function. Set the valve failure critical threshold. =2.0 (determined through historical failure data). By solving this equation, the time probability distribution of the composite health index reaching the failure threshold of 2.0 for the first time from the current value of 1.58 is obtained.

[0162] Failure probability and remaining lifetime calculation, time-varying failure probability: Calculate the probability that the index will exceed 2.0 within the next 72 hours: 24-hour failure probability: 18%; 48-hour failure probability: 65%; 72-hour failure probability: 92%.

[0163] Average remaining useful lifetime: Based on the first pass time distribution calculation, the average remaining lifetime is 42 hours.

[0164] Control signal mapping: First analog control signal (probability dashboard): Maps the 72-hour failure probability of 92% to the dashboard pointer deflection angle (0° corresponds to 0%, 270° corresponds to 100%), driving the pointer to deflect to the 248° position. Second analog control signal (remaining life progress bar): Maps 42 hours to the progress bar display length (full scale 100 hours), controlling the progress bar to display 42% of the length.

[0165] 206. Composite Module for Multidimensional Health State Space Construction and Visualization Mapping Steps: Using entropy-producing health indicators and chaotic indicators as coordinate axes, construct a two-dimensional health state phase plane to comprehensively characterize the evolution of valve health state; on this two-dimensional health state phase plane, define an elliptical or custom-bounded stable health region based on the cluster distribution of indicator data under the initial health state of the valve; dynamically adjust the boundary of the stable health region based on the time-varying failure probability, where the width of the boundary region is proportional to the time-varying failure probability; map the instantaneous value of the composite health indicator to the visual warning attribute of the corresponding coordinate point on the two-dimensional health state phase plane; the visual warning attribute includes the color and size of the point, where the color of the point changes continuously on a spectrum from green to red according to the time-varying failure probability, and the size of the point is proportional to the degree of deviation of the composite health indicator from its health benchmark value; Warning Signal Generation and Output Steps: When the coordinate point continuously deviates from the center of the stable health region and its corresponding time-varying failure probability... When the first preset threshold is exceeded, a first-level warning signal is generated; when the coordinate point touches or crosses the dynamically adjusted boundary of the healthy and stable region, and its corresponding time-varying failure probability exceeds a higher second preset threshold, a second-level warning signal is generated; the first-level and second-level warning signals are converted into standard industrial communication protocol signals for output; the fault tracing auxiliary steps based on physical mechanisms are as follows: after generating the second-level warning signal, the direction of the coordinate point's motion trajectory on the two-dimensional healthy state phase plane is analyzed; if the trajectory mainly extends along the entropy production health index axis, an auxiliary judgment signal indicating abnormal flow channels (wear, corrosion, scaling) is generated; if the trajectory mainly extends along the chaotic index axis, an auxiliary judgment signal indicating abnormal structural strength (crack initiation, loose connectors) is generated; if the trajectory extends along both axes simultaneously, an auxiliary judgment signal indicating a composite fault mode is generated; the auxiliary judgment signal and the warning signal are output synchronously to provide a physical mechanism reference for the fault type for maintenance decisions.

[0166] It should be noted that the high-pressure regulating valve on the main steam pipeline of a 600MW unit in a thermal power plant is used as an example to illustrate the process of constructing and implementing a multi-dimensional health state space. Assume the current monitoring data are: entropy production health index 1.15 (thermodynamic anomaly), chaos index 3.75 (kinetic instability), composite health index 1.58 (comprehensive degradation), and time-varying failure probability 92% (within 72 hours).

[0167] A two-dimensional health state phase plane is constructed with entropy production health index as the horizontal axis (range 0-3) and chaos index as the vertical axis (range 0-5).

[0168] Based on the initial health state data cluster of valves (100 sets of baseline data), an elliptical healthy and stable region is defined: center point coordinates (1.0, 1.0); major axis radius 0.15 (entropy production direction); minor axis radius 0.12 (chaotic direction);

[0169] Dynamic boundary adjustment and visual mapping: The boundary of the healthy area is dynamically adjusted according to the time-varying failure probability of 92%: the boundary width is expanded to 3.2 times the initial value (calculation coefficient: 1 + 0.024 × 92 ≈ 3.2).

[0170] The current status point coordinates (1.15, 3.75) are mapped as follows: Color: Dark red (RGB: 180, 20, 30) based on a 92% failure probability; Size: Diameter 8mm (proportional to the deviation of the composite health index 1.58 - 1.0 = 0.58).

[0171] A warning signal is generated when the state point (1.15, 3.75) continues to deviate from the center (1.0, 1.0) and the failure probability is 92% > 80% (second threshold), triggering a second-level warning signal.

[0172] Output standard industrial signal: 4-20mA; analog signal: 16.8mA (corresponding to 92% probability); Modbus TCP protocol transmission warning code: 0x02 (level 2 warning);

[0173] Fault tracing auxiliary analysis, analyzing the motion trajectory of state points (trajectory directions of the 10 most recent points):

[0174] The change in entropy production index ΔS = +0.18 (an increase of 18.5% compared to before 10:00).

[0175] The chaos index changed by ΔC = +1.25 (a 50% increase compared to before 10:00).

[0176] The trajectory direction angle θ = arctan(1.25 / 0.18) ≈ 81.7° (close to the chaotic axis);

[0177] Generate structural strength abnormality auxiliary signal: Output fault type code: 0x12 (indicating crack initiation / loose connection); Synchronously output maintenance suggestion: "It is recommended to check the tightness of the valve stem connection and the cracks on the inner wall of the valve body";

[0178] This visualization system intuitively presents the valve health degradation process through real-time changes in coordinate point position, color, and size, providing maintenance personnel with decision support that combines physical mechanisms and quantitative assessments.

[0179] In this embodiment of the invention, a distributed acquisition system with a unified time base and the PTP1588 precision clock protocol are used to achieve microsecond-level time synchronization, ensuring strict time alignment of pressure, temperature, flow, and vibration data, eliminating asynchronous measurement errors, providing a high-precision data foundation for subsequent thermodynamic and kinetic analysis, and improving the accuracy of anomaly detection. Based on the second law of thermodynamics, a dimensionless entropy production health index is generated by calculating the ratio of the absolute entropy production rate to the minimum entropy production rate of the health benchmark. This quantitatively characterizes the degree to which the energy dissipation inside the valve deviates from the health benchmark, enabling early detection of faults such as wear or scaling in the flow channel and preventing performance degradation. A dimensionless chaos index is generated by using phase space reconstruction and maximum Lyapunov exponent calculation, combined with nonlinear time series analysis technology. This quantifies the sensitivity of the system's dynamic trajectory to initial conditions, enabling early detection of structural instability problems such as valve core wear and loose connections. A weighting factor is dynamically calculated, integrating the entropy production health index with... Chaotic indicators generate more sensitive composite health indicators that comprehensively reflect the overall health status of valves, amplify abnormal signals, and trigger early warnings earlier than single indicators, improving the timeliness of fault detection. Based on the time series of composite health indicators, a stochastic dynamic system evolution model is constructed to solve the Fokker-Planck equation, obtain the future time-varying failure probability, achieve advanced early warning, quantify remaining lifespan, provide a scientific basis for operation and maintenance decisions, and reduce unplanned downtime. Using entropy-generated health indicators and chaotic indicators as coordinate axes, a two-dimensional phase plane is constructed to dynamically adjust the boundary of the healthy and stable region, map visual warning attributes, intuitively present the valve health status degradation process, and reflect failure risks in real time through color and size changes, improving operation and maintenance efficiency. Analyzing the direction of the motion trajectory of state points generates auxiliary judgment signals indicating abnormalities in the flow channel or structural strength, providing a physical mechanism reference for fault types for maintenance decisions, reducing misdiagnosis, and optimizing maintenance strategies.

[0180] Figure 3 This is a schematic diagram of a high-pressure valve flow capacity anomaly detection and strength failure early warning device provided in an embodiment of the present invention. The high-pressure valve flow capacity anomaly detection and strength failure early warning device 300 can vary considerably due to different configurations or performance. The device 300 includes a transmitter 301, a receiver 302, and a processor 303. The processor 303 can also be a controller. Figure 3 The device is designated as "controller / processor 303". Optionally, the device 300 may also include a modem processor 305, which may include an encoder 306, a modulator 307, a decoder 308, and a demodulator 309.

[0181] In one example, transmitter 301 modulates (e.g., analog-to-analog conversion, filtering, amplification, and up-conversion, etc.) the output sample and generates an uplink signal, which is transmitted via an antenna to an access network device. On the downlink, the antenna receives the downlink signal transmitted by the access network device. Receiver 302 modulates (e.g., filtering, amplification, down-conversion, and digitization, etc.) the signal received from the antenna and provides an input sample. In modem processor 305, encoder 306 receives service data and signaling messages to be transmitted on the uplink and processes (e.g., formatting, encoding, and interleaving) the service data and signaling messages. Modulator 307 further processes (e.g., symbol mapping and modulation) the encoded service data and signaling messages and provides an output sample. Demodulator 309 processes (e.g., demodulates) the input sample and provides a symbol estimate. Decoder 308 processes (e.g., deinterleaving and decoding) the symbol estimate and provides decoded data and signaling messages to device 300. Encoder 306, modulator 307, demodulator 309, and decoder 308 can be implemented by a combined modem processor 305. These units perform processing according to the radio access technology adopted by the radio access network (e.g., LTE and other evolved systems access technologies). It should be noted that when device 300 does not include modem processor 305, the above-mentioned functions of modem processor 305 can also be performed by processor 303.

[0182] The processor 303 controls and manages the operation of the device 300, and is used to execute the processing procedures performed by the device 300 in the above embodiments of this disclosure. For example, the processor 303 is also used to execute various steps of the transmitting or receiving device in the above method embodiments, and / or other steps of the technical solutions described in the embodiments of this disclosure.

[0183] Furthermore, the device 300 may also include a memory 304 for storing program code and data for the device 300.

[0184] Understandable Figure 3 Only a simplified design of device 300 is shown. In practical applications, device 300 can include any number of transmitters, receivers, processors, modem processors, memory, etc., and all devices that can implement the embodiments of this disclosure are within the protection scope of the embodiments of this disclosure.

[0185] The present invention also provides a high-pressure valve flow capacity abnormality detection and strength failure early warning device. The high-pressure valve flow capacity abnormality detection and strength failure early warning device includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor performs the steps of the high-pressure valve flow capacity abnormality detection and strength failure early warning system in the above embodiments.

[0186] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the high-pressure valve flow capacity abnormality detection and strength failure early warning system.

[0187] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0188] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0189] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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. Such 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.

Claims

1. A high-pressure valve flow capacity abnormality detection and strength failure early warning system, characterized in that, The high-pressure valve flow capacity abnormality detection and strength failure early warning system includes: The acquisition module is used to simultaneously measure the fluid pressure and temperature at the inlet and outlet of the high-pressure valve, the fluid mass flow rate, and the vibration acceleration of the valve body. The processing module is used to obtain entropy production health indicators based on the acquired fluid pressure, temperature, and fluid mass flow rate, according to the laws of thermodynamics, including: The first specific entropy value of the fluid at the valve inlet is determined based on the pressure and temperature at the inlet, and the second specific entropy value of the fluid at the valve outlet is determined based on the pressure and temperature at the outlet. The absolute entropy production rate of the valve under the current operating state is calculated based on the fluid mass flow rate, the first specific entropy value, and the second specific entropy value. The absolute entropy production rate is then compared with the preset minimum entropy production rate of the valve under a healthy baseline state to generate an entropy production health index. The configuration module is used to obtain chaotic indices characterizing the degree of instability in the dynamic state of the valve structure based on the vibration acceleration and using nonlinear time series analysis techniques, including: A multidimensional phase space trajectory is generated by reconstructing the phase space of the vibration acceleration signal sequence. The long-term evolution behavior of neighboring points on the phase space trajectory is tracked in the reconstructed multidimensional phase space to obtain the maximum Lyapunov exponent. The maximum Lyapunov exponent is compared with the reference Lyapunov exponent benchmark value that characterizes the background noise level of the system to obtain a chaotic index used to characterize the degree of instability of the dynamic state of the valve structure. The fusion module is used to fuse the entropy production health index and the chaos index to generate a composite health index, and to use the change sequence of the composite health index over time as the system state evolution characteristic trajectory. The allocation module is used to construct a stochastic dynamic system evolution model based on the system state evolution characteristic trajectory, and project it to obtain the failure probability of the valve's future state, so as to achieve early warning of abnormal flow capacity and strength failure risk.

2. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 1, characterized in that, include: The system collects the first pressure and first temperature measurements of the fluid at the inlet of the high-pressure valve, the second pressure and second temperature measurements of the fluid at the outlet of the high-pressure valve, the real-time mass flow rate of the fluid flowing through the high-pressure valve, and the simulated vibration acceleration signal of the valve body surface of the high-pressure valve. The analog vibration acceleration signal is converted from analog to digital to generate a sequence of digital vibration acceleration signals sampled at equal time intervals. For each data point in the first pressure measurement value, the first temperature measurement value, the second pressure measurement value, the second temperature measurement value, the real-time mass flow rate measurement value, and the vibration acceleration digital signal sequence, a timestamp is generated to form a multi-parameter synchronous measurement dataset.

3. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 1, characterized in that, Set the actual entropy production as : ； in, For the import specific entropy, For export entropy, For mass flow rate; ； This is an indicator of entropy production health.

4. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 1, characterized in that, include: Based on the statistical dispersion characteristics of the real-time time series of the entropy production health index and the real-time time series of the chaos index, the real-time weighting factor is dynamically calculated. The instantaneous values ​​of the real-time weighting factor, the entropy-producing health index, and the chaos index are input into a preset nonlinear function to generate an instantaneous composite health index value. The instantaneous composite health index values ​​are collected sequentially over time to form a composite health index time series trajectory that characterizes the degradation path of valve health status.

5. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 4, characterized in that, include: Based on the time series trajectory of composite health indicators, the drift coefficient and diffusion coefficient of stochastic differential equations are identified online by calculating their conditional statistical moments under different time delays. Based on the identified drift coefficient and diffusion coefficient, a Fokker-Planck equation is constructed to describe the evolution of the probability density function of composite health indicators over time. A composite health index threshold corresponding to the critical state of valve failure is set, and the probability density distribution of the composite health index reaching the failure threshold for the first time from the current value is obtained by solving the Fokker-Planck equation, so as to obtain the time-varying failure probability and the average remaining useful life estimate within a specified time period in the future. The time-varying failure probability is mapped to a first analog control signal for controlling the deflection angle of a probability dashboard pointer, and the average remaining useful life estimate is mapped to a second analog control signal for controlling the display length of a remaining useful life progress bar.

6. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 5, characterized in that, It also includes composite modules: Using entropy production health indicators and chaos indicators as coordinate axes, a two-dimensional health state phase plane is constructed. On this two-dimensional health state phase plane, a healthy and stable region is defined based on the cluster distribution of indicator data under the initial health state of the valve. The boundary of the healthy and stable region is dynamically adjusted based on the time-varying failure probability, where the width of the boundary region is proportional to the time-varying failure probability. The instantaneous value of the composite health index is mapped to the visual warning attribute of the corresponding coordinate point on the two-dimensional health state phase plane.

7. The high-pressure valve flow capacity abnormality detection and strength failure early warning system according to claim 6, characterized in that, When the coordinate point continuously deviates from the center of the healthy and stable region and its corresponding time-varying failure probability exceeds the first preset threshold, a first-level warning signal is generated. When the coordinate point touches or crosses the dynamically adjusted boundary of the healthy and stable region and its corresponding time-varying failure probability exceeds the second preset threshold, a second-level warning signal is generated. The first-level warning signal and the second-level warning signal are converted into industrial communication protocol signals for output. After generating the second-level early warning signal, the direction of the motion trajectory of the coordinate point on the two-dimensional health state phase plane is analyzed. If the trajectory mainly extends along the entropy production health index axis, an auxiliary judgment signal indicating abnormal flow channel is generated. If the trajectory mainly extends along the chaos index axis, an auxiliary judgment signal indicating abnormal structural strength is generated. If the trajectory extends along both axes simultaneously, an auxiliary judgment signal indicating a composite fault mode is generated. The auxiliary judgment signal and the early warning signal are output synchronously.

Images